Speed as a system
The complete NFX course on startup speed

How to use this file

This is a textbook, not an article and not meeting notes. It is structured so that you can study it sequentially, like a course book. The file contains nine modules, one for each masterclass episode, and four cross-cutting sections at the end.

Each module follows the same template. First come the learning objectives: what exactly you should be able to do after the module. Then the theory unfolds: every concept of the episode is explained in full, in simple sentences, with definitions and logic. Within the theory there are real-world examples with exact numbers and the names of companies, people and products, because in this course it is the numbers and the cases that carry the evidentiary weight. Next comes the Key idea callout: the one idea the whole module exists for. Then a row of cards Numbers to remember: a set of figures and values to keep in mind. After that, the Common mistakes section: the typical ways a founder breaks what the module describes. And at the end, Self-check questions: five to eight questions to answer from memory, without looking at the text.

If you are studying the course for the first time, go from module one to module nine: the modules are built so that each one builds on the previous one. If you are already familiar with the material and want a quick refresher on a specific topic, use the course map below and jump straight to the module you need. The diagrams in the file serve as mnemonics: each can be recalled separately from the text and used to reconstruct the logic of the section.

An important note on the source. All the material comes from an oral course, that is, it is the self-report of the author and his firm, not an independently verified study. The numbers in the examples are often rounded in speech, some stories are told from memory and in places diverge between episodes. At the end of the file there is a separate section that honestly lists these divergences and caveats. The educational value is not lost because of this, but a critical attitude toward the source should be maintained throughout the reading.

Course map

Below is the whole course at a glance. Nine modules correspond to nine episodes, four cross-cutting sections complete the file.

The four levers the course lays out across all the episodes are speed, language, idea and psychology. The nine modules divide the material by episode, but cross-cutting section A gathers them back into a single system.

Module 1. The Speed Talk: speed as the only criterion

Episode 1 of 9 · course foundation

Why speed is the only answer to the question of the right founder

When the author is asked what exactly tells him that he is dealing with the right founders, the answer is one: speed. Not experience, not market size, not charisma, but speed itself. It shows up in the small things. Whether a person answers an email quickly, whether they change a marketing plan quickly, whether they fix their spreadsheets quickly. Speed is felt physically: the person is uninhibited and simply moves forward instead of dragging along. That is the marker, because behind behavioral speed lies an absence of internal brakes, and an absence of brakes is what distinguishes a team capable of outrunning a larger opponent.

The course's key metaphor is taken from Star Wars. The founder is an X-Wing fighter pilot going up against the empire. What matters is that the times have changed: this is not 1994, when there were only a few big players, like Microsoft and AOL. Now the founder faces numerous Death Stars, that is, large established corporations. The only advantage a small ship has against a huge station is full throttle. The hero enters the attack saying that he is going in at full power. His best friend Wedge flies beside him, and to the order to go at full throttle he answers with one line: "I am with you, boss." He does not ask whether this is the right decision, does not offer to discuss it, does not want to make sure first. And it is Wedge who is the only one to survive that battle. The difference between him and those who asked questions is the difference between speed and caution, and in the heat of the fight speed wins.

From the metaphor follows an arithmetic conclusion. If you move twice as fast, you can outrun anyone bigger or smarter than you, because you make two moves for every one of theirs. You can make three, five, twenty moves for one of theirs. That is your advantage. Hence the rule: you must not merely move fast, you must think in terms of speed.

Five benefits of speed

Speed delivers measurable benefits, and they are worth listing separately, because they explain why it is worth enduring the discomfort of fast movement.

Why the speed bar is set low: school and corporation

The main conclusion of the section on obstacles is this: almost everything that keeps you from moving fast is in your head, not in the world. The speed bar is low for most founders because it was lowered by the two institutions almost every successful founder passed through.

The first institution is school. How fast does school work? It has calendars, semesters, long cycles. School is extremely slow, yet it is where the person achieved success. Someone who got into Stanford was the best at school, then was good at a corporation. And how fast does a corporation work? Also not fast, and that does not suit it. Moving too fast frightens people, people need to be informed, they need to exercise judgment, they need things tweaked and approved. So the successful founder was badly trained by his environment and did not notice it. It seemed to him that he learned at school or at a job at Google and Microsoft, but in fact he learned how to move too slowly.

This is not the founder's fault. People who were not so successful move faster, because they have nothing to lose. A successful person feels that he has something to lose, and feels that he understands how the world works, since he has already achieved success. But the mistake was made not by him but by his environment, and it can be escaped. That is exactly why the course calls its first part a reset of the speed bar: you need to feel in your heart that a fast pace is possible and will not be painful.

Mike Cassidy: the father of speed

The course names Mike Cassidy as the founder of the speed movement, who lives in Los Altos and continues to work. In the nineties he broke what had not been broken before him. The standard startup launch cycle back then looked like this: six months to research the idea, three months to raise money, three months to hire the core team, thirteen months to develop the product and three months to launch. That is about twenty-four months in total, and it was considered a brilliant result, something like "startupland, you are crazy."

Cassidy made the whole journey in four months. He raised the money literally in two days: he packed all the meetings into one day and at the end of each meeting told the investor that he had six more meetings that day, that one term sheet was already in hand, and that if they wanted in they should let him know that evening, here were the terms. He stated plainly the price and the amount at which an investor could win the deal. By the end of the day he had two or three offers, and the next day or two days later he announced the winner.

He hired the team in two weeks. At the end of the interview he already had a contract ready to sign, with the price and all the terms, and he told the candidate: you have twenty-four hours, either you sign and start tomorrow, or you do not. He built the product very fast, and it was he who coached the course author in the late nineties at his first startup: it felt like the presence of Obi-Wan. One of his questions, which was remembered forever, was simple: why is this taking you so long?

The author's own experience: an incubator and twenty products

In 2007 the author and his team started an incubator. They had an exit from their first company and some money. Forty-five employees sat in one room and built four products at once. The feeling was happy and joyful. They managed to build twenty products for a tenth of the money competitors spent and a tenth of the time, precisely thanks to the speed tactics. This is personal experience, not only someone else's, and the course needs it to prove that what is described is really achievable.

Twenty-one days and thirty-five days: a game development story

One of the strongest calibration examples came from the game business. The company hired people from the game industry, and they said that a good game you could be proud of would take twelve or fourteen months. The author replied that the team had about thirty-five days. The people looked down and sulked: that is impossible. And then one web engineer, who had never worked in the game industry and so did not know it could not be done that way, said: I will do it in twenty-one days.

He was asked what he needed. The answer was: no meetings, just let me build this thing and give me half a designer instead of a whole one, because a whole designer would take up his time. In twenty-one days he delivered a working game. Three weeks after that the game had three million users. The game specialists were stunned: he did it because he did not know it could not be done that way. He promised, worked all weekends and all nights, had fun, and the game brought the company a lot of revenue. Hence the rule: impossibility often exists only in the head of someone who already knows that this is not "how it is done."

Dragons of Atlantis: five months against twelve

Another game from the same company was Dragons of Atlantis. When the company had only 600,000 dollars left and was about to shut down, the specialists said that a serious game of that scale would need four million dollars and twelve months. There was neither the money nor the time. The team gathered and built it in five months for four hundred and fifty thousand dollars, and the next year it did one hundred and twenty million dollars in revenue.

The game had forty-five million registered users and at one point three million concurrent users, who spent two to eight hours a day in the game. The product was shipped four times a day. Hence the ironclad argument against engineers who say they cannot ship updates once a day or once a week: they do not have three million concurrent users, so they can do more. Most firms ship something once a week or once every two weeks. At the author's accelerator it took weeks to convince founders that daily releases were even possible, because they had been trained at Microsoft or Google. But when the convincing worked and they started shipping every day, they caught their breath and the acceleration began.

The author's son and the solo Atlantic crossing

In January of one year the author's son did not know how to sail and asked himself whether he could cross the Atlantic alone. Usually people prepare for this for five years and spend three hundred and fifty thousand dollars. This teenager started learning to sail in March, set out in June, crossed the Atlantic in a boat for twelve thousand dollars and finished at the end of July. He got on the Today show, got articles in Cruising World magazine and awards at the New York Yacht Club. Seven months and twenty thousand dollars against the industry's five years and three hundred and fifty thousand.

The network of people who spent three hundred and fifty thousand hates such a result. They call it dangerous, wrong, almost evil. In fact they simply have a low speed bar. The son meanwhile had a great time: that summer he planned to cross the Mediterranean again, already seventeen years old. The conclusion the course draws from this story is this: do not think about age or experience, do not believe how much time and money everything is supposed to take. That is not true, they are all mistaken.

The rule of experimentation: twenty to a hundred a month

How often do you experiment with the product? With hardware it is harder, which is why they say "hardware is hard." But if you make software, you should run between twenty and a hundred experiments a month. In the case of Dragons of Atlantis the team ran twenty to a hundred experiments a week. If you cannot do twenty to a hundred a month, you are simply not in the game. An experiment is the smallest single act of change that can be shipped and measured; the frequency of experiments is the frequency of cycles, and it is exactly what shapes the speed of the company.

The home list: Hot or Not, Jiff, Direct Hit, Xfire

There is a set of cases the course cites as an evidentiary list proving that compressing timelines is real. Mike Cassidy built Direct Hit in three months and sold it for five hundred million dollars in five hundred days. He also built Xfire in three months and sold it to MTV in roughly twelve months for one hundred and ten million dollars. James Hong built Hot or Not in eight days, and his company was called "Eight Days": he decided not to choose what to build but to think about what could be done in eight days, and if something did not fit in eight days, then it did not get built. The author also founded a company called Jiff, when the money ran out again. He had a presentation coming up in thirty days, and there was no product. He designed, built and launched the product in thirty days, gave the presentation and raised seven and a half million dollars when there were only thirty days of capital left. The team was told not "what we want to build" but "we have thirty days, here is the presentation, so we need to design, build and launch in thirty days." It was an external force, not a boss's shouting, that helped people understand that no one was tormenting them and there was no need to suffer, they just had to get the job done.

Meerkat and Clubhouse were thirty-day projects. The carousel of examples should convince the reader that the speed bar can be raised very high, and if someone says it is impossible, you do not have to believe them.

The main obstacle is fear, and twelve ways to get around it

The course openly admits: the main barrier is fear. It is fear that makes everyone slow. Then twelve ways to get around fear are given, and this is the core of the first module. Each way is a separate little mental technique that you can apply literally today.

1. The product is not you

A voice in your head says: "This is my soul, my heart, this is me, everyone will love the product and love me, and then I will be enough, then they will love me." That is not so. The product is not you, it is just a product, a set of pixels, a set of features, the start of the journey. You just have to send it out the door. The author gives the example of someone asking whether Mark Zuckerberg should spend time on photo sharing in college. The point is that even the greats started with something small and did not try to get everything right at once.

2. Do not chase correctness, you need step-by-step development

This applies especially to "product" people, who always want to do everything right and perfectly. The author himself is a product person and says this with irony. Hiring a VP of Product often slows things down. He personally does not hire a VP of Product: there is an engineer, there is a designer, there is a CEO, and that is enough. Do not try to do it right, do it gradually, step by step. First the product needs to be shipped, then it needs to be measured.

Here an important proportion is given: half of the product is the product, the other half is metrics. More precisely, roughly ten percent is the product, fifty percent is metrics and forty percent is the onboarding flow. That is, the real work on the product consists mostly of measurement and of how a person enters the product.

3. A skateboard, not a convertible

This is one of the most famous diagrams of the course, one that founders keep coming back to for years. Most people think like this: first the wheel, then the chassis, then the body, and at the end you get a beautiful convertible that is a pleasure to drive. But products are not built that way. You have to start with a skateboard, which has nothing at all to do with a car. You have a vision of what you will be proud of at the end, but you cannot jump to the end. You have to start with a skateboard, fall in love with the skateboard, make the skateboard excellent. Then the skateboard turns into a scooter, a kick scooter, not electric, then a bicycle, then a motorcycle, and only then a car. The founder who starts with a skateboard looks like a loser when everyone around is riding bicycles. But he is the one who will reach the car fastest, because he already has a skateboard and is actually moving.

4. Launch behind the scenes, on a different URL

If the company is called NFX, do not run the experiment on nfx.com, run it on another domain. Find a small group of people and show it to them. One of the author's teams had a huge problem: they were so afraid for the brand that they could not launch anything new. So they created many different URLs for them, so they would feel safe. A person tells a friend that he works at NFX, the friend goes to the NFX site and says it is cool, while everything broken lives on other sites the friends do not see. The team relaxes, the fear goes away, and you can move fast.

5. These are experiments, not a business

Many founders of the last decade "play at business." They saw "The Social Network," read too many blogs, tech founders became idols and rock stars, and everyone wants to be at the end, in the sports car. They spend months thinking about how to properly run a board meeting instead of just running it. The course says: do not turn business into a performance. Your job is a series of experiments, not a single business. "Give me two million, I will run a series of experiments." A series of experiments looks like a skateboard, and that is the state of mind that lowers fear.

6. The studio model: one idea every thirty days

The Meerkat story, which the author promised to tell, shows the studio model. A team from Israel built a product called Yevvo for fourteen months, and it did not work. The author told them to try another product with seven specific features. After eight months they still did not have one or two of the features, and it was clear it would not work. They were asked to test it on twenty-five and on a hundred people, and it became obvious it would not work. Finally they reached the limit, their own estimate that quality takes eight months broke down.

Then they stood at a small whiteboard and in ninety minutes came up with five ideas. Each was given its own color, its own name, one or two features, and they decided to launch all five, one every thirty days. The team said it was impossible. The author replied that it was possible, and explained the logic: if you know the blue idea will come out thirty days after the yellow one, you will not worry too much about yellow, because you know the next one is already coming. As you move toward red or pink, something workable may emerge. In reality Meerkat shipped in sixty days, but that was still far better than eight months and the fourteen months before that. So the path compressed from fourteen to eight to two months. Meerkat took off, gained users and press, but then Twitter cut it off. Yet the next product, Houseparty, was built in thirty days and became big. The company was eventually sold, and the green, red and pink ideas were never used, but they are always ready. The moral: think of yourself as a studio and simply get, through experimentation, to what works.

7. Your job is to learn

Your job is to run experiments and learn. If you are given two million and asked what you are busy with today, the right answer is: learning. And the next question: what did you learn today? That is exactly what the conversation is about.

8. The borrowing mindset

Another way to get around fear is to borrow, to copy what already works for others. Do not argue about it, just take what works. Need a calendar, take the best calendar, copy something. That is fine, it will still come out as your own, because you have a unique mind. The author tells a story: he acquired one person's first company, that person worked for him for a year, then together they came up with a new idea, the author gave him the URL and said to let him know if it worked. Seventy-two hours later the man wrote that it worked. The author asked for the code. That person's company was called Bebo. The code was sent and implemented at his company in seven days, because at a company of a hundred people it was too slow. Both grew on the very same product, on the very same code, because they had come up with the idea and the design together. The former employee grew fifty-three million users, and the author's company forty-seven million in six months: that is how viral such products were. In the end the author's company was acquired by Monster and remained a photo-sharing service, while the former employee turned his product into a social network and sold it for eight hundred and fifty million dollars. Although they started from absolutely identical code, the difference in situation and in people brought them to different places. The conclusion: do not fear that you will not create anything unique, even if you copy for a year, the result will be unique because of who you are, whom you talk to and where you are.

Reinforcement of this argument: Google invented nothing, it copied everything. The author once said that Google copied everything except AdSense, but the founder of AdSense raised his hand and said that his was simply bought, so Google copied absolutely everything. Facebook copied Friendster. Airbnb is VRBO with a ten percent model instead of three hundred dollars a month, and it started as Rent-My-Room, and took off when it came to look more like VRBO, which had existed since 1995. Uber did not invent Uber; thirty days after Lyft launched, it copied it, pulling everyone off other tasks and putting them on UberX. Lyft did not invent it either: Sidecar invented it, and no one heard of Sidecar because it was too slow. Lyft had twenty-eight people, twenty-six of them put on Lyft (the company was originally called Zimmeride), and it took off. Uber copied it thirty days later, raised more money, became more aggressive and grew six to eight times bigger than Lyft, while Sidecar died. So do not be afraid to borrow, because everyone does it.

9. The "screw it" mindset

Most of the best things in the author's business career happened because "screw it" was said. The money was running out, his back was against the wall, he suffered, did not sleep, and at that moment he gave in and did the right thing, saving the business. He was getting in his own way, he overcame his own fear and did what the market wanted. He could have seen it earlier. So do not repeat his mistake: do not wait until there are thirty or sixty days of capital left to reach the "screw it" moment. Start that way from the very beginning, and it will be much more fun.

10. The chip on the shoulder and self-sabotage

Many found companies to prove something. A person may have a chip on his shoulder, a bad relationship with his father, it is very common. The author asked one of Elon Musk's friends whether it was noticeable, while they were living together in Los Angeles, that he was the greatest entrepreneur in the world. He answered no, he was like everyone else. Twenty minutes later he came back and said there was one thing after all: what helped him was that his father told him he would never achieve anything. This is Silicon Valley's best venture capitalist, George Zachary, and that was his explanation of what made Musk Musk. Outstanding founders have psychological problems, something drives them, something is a little off. But if you are trying to prove something, part of you may believe you are unworthy, and then, without noticing, you engage in self-sabotage. You do not sleep, you try as hard as you can, you hire people, and everything external seems to prove that you are not destroying yourself but moving forward. But practically everyone does this to some degree, and you have to be aware of it and point it out to your co-founders. Hence: Google "self-sabotage," read books, watch short videos, that is enough to get a handle on this state. When the author lines up his founders on a scale from speed to success, the correlation is very clear, and the psychology of the slow ones mostly revolves around self-sabotage: they go slowly to prove to themselves that they are failures, although they founded a company to prove the opposite.

11. Stop being a pushover

A founder often gets in his own way. You just have to make a decision. There is a choice. If it is not working, you can drop the business and admit defeat. You can change hesitantly, keeping the old product and launching the new one at the same time, that is, take half-measures, and that is exactly self-sabotage and confusion. Or, when you realize it is not working, you can move the whole company to the new thing and do it today. When the author sees founders who do that, he guarantees they will be more successful than those who take half-measures and hedge.

12. Do it for the people who count on you

If you do not want to help yourself and do not want to move fast for your own sake, do it for others. You have employees, teammates, investors who count on you. You owe them. This is another way to get out of your own head: to think about everyone who depends on you.

The list of simplifications: what to simplify in order to speed up

After dealing with fear, the course gives a long list of simplifications. The point is that most complications give nothing but delay, and they can simply be removed.

It is worth recalling separately the marketing budget and the board of directors. When the author sat on the Second Life board, they spent twenty-five minutes discussing how to spend ten thousand dollars a month on marketing, and he noted that the discussion itself had already cost fifty thousand. The conclusion: the budget should head toward zero, because the fastest path is to let the product grow on its own. By the same logic the board of directors is kept to three people. One giant ten-billion-dollar company has only three people on its board, and one of them is from Tesla, and that is enough; the author had a board larger than three at only one startup, and it was a total mess.

A separate word on term sheet terms. East Coast investors often offer warrants, milestones, contingencies, rights of first, participation, all on their terms. You must not do this. The author was five minutes late to a meeting precisely because he was arguing with founders he had warned about this for two and a half years, and they did not listen. Now they have an unpleasant person on the board who is digging into the company and intends to take it over and cheat the founders. Rule: keep it vanilla. If an investor says "I will give you a twenty-eight-million valuation if you allow these contingencies," answer: "I will take an eighteen-million valuation, but let everything be vanilla." You will make more on that, move faster and have more fun.

Warren Adams, the author of the first social service on the internet in 1995, handed the author a check for one hundred and twenty-five thousand dollars and then pulled it back and said: promise me one thing, that you will not do partnerships. At that time the internet was built through partnerships with AOL, MSN and the like. The author promised, took the check and still went off to try partnerships, but Adams's voice rang in his head, and within four weeks, instead of four to eight months, it became clear where it was heading: endless negotiation, endless integration, endless metrics, divergent interests. The rule is simple: you cannot build a big business on biz-dev, it is a complication and playing at business, not movement.

Go where the movement is, and find the fast water

The biggest waste of time and the biggest source of slowness is building what people do not want. So you have to do research: what already existed? This is not 1994, this is not a blank slate, someone has already tried your idea or something like it. Research will save months and years. The author tells his founders two or three times a week: go on LinkedIn, find a VP of Sales who worked at your competitor, invite him to lunch and ask questions. They answer that he will not tell them anything. They are told: try. Then after meetings with six people they come back and say their life has been turned upside down. This is a very good tactic, because software is already old, and there are plenty of people around with industry knowledge.

The second conclusion: you have to find the fast water. If things are going okay, you are in the river, but you are asleep, playing at business and growing ten percent a month. That is no good: you need to grow seventy percent a month. The one growing ten percent is sitting in slow water and thinks everything is fine only because the board has not fired him yet.

10%per month: slow water, self-complacency 70%per month: fast water, where you need to be

Attract resources: the Zynga and Texas HoldEm Poker case

There was a period when a great many people were trying to get onto the Facebook platform. Everyone knew the deal terms, but one man, at the head of Zynga, attracted resources faster than anyone. In literally two weeks he bought an existing game, Texas HoldEm Poker, that is, he borrowed what already worked. Then he raised thirty-five million dollars and within two months hired about one hundred and fifty people. These were not the best people, many of them could not find work elsewhere, which is why they were available to hire in the next sixty days. But he had a lot of people, a lot of money, a lot of press and a lot of traffic, he bought traffic and did remarketing to these people, and so he spun up a business with network effects around games on the Facebook platform. At one point his company was worth eleven billion dollars. He was more successful than anyone on that platform during the gold rush. The conclusion: you have to pull all these nodes into your company's network, and the faster you do it, the faster you grow.

Hiring as an art and team size

Hiring must be turned into an art. The offer is made before the candidate leaves, and filling a position gets a maximum of three weeks. Do not let yourself be convinced it will take six months. New AI tools in hiring only help. About the team the author says this: you need the mindset of "I came to kick ass and take names," and to get your own ass kicked by the market, by investors, by co-founders. This should be welcomed, because if you are not allowed to get hit, you will not make changes fast enough. You have to be a competitive person, people should say you are quite competitive, even if it sounds disapproving. Being a founder is a gladiator sport, not a career path. It is a prison until you exit the company. Employees, investors, your reputation all count on you.

An example of the savage: one of the founders of Twist Bioscience is Emily LeProust, a Frenchwoman, she did patents, raised money, did sales, hired everyone, she is a monster entrepreneur. She likes to say that she likes a person who is "sauvage," that is, "wild" in French. She looks for wild founders, but not in the sense of mean: she means that they are decisive, aggressive, make decisions fast, possess speed.

Team size is a problem, and you must keep the team minimal. One person talks to a second, a third arrives, and now everyone has to talk to two others, say good morning, and imperceptibly a crowd of people appears around the table, and everything moves very slowly. Then it starts: "I was not invited to the meeting," "I thought I would make the decision about this," "love me, let me belong, listen to what I have to say, let me be creative." All human needs start getting in the way of speed. You must understand clearly that four right engineers are better than six marketers, and one right marketer is better than six, and thirty-five engineers is slower and worse than a smaller number of the right ones. The author fired his first three VPs of Product. The last one said to him: cheap, fast or good, pick two. It sounds smart and profound, but it is nonsense. When you find something cheap and fast, it usually becomes good, because you can iterate it to good. Cheap, fast and good go together, and that is usually how it is. After that the author never hired another VP of Product.

Stoke: the math of eighty thousand dollars a day

One of the author's portfolio companies in Seattle builds metal rockets for spaceflight. Its CEO is an engineer and runs the company by math. The company's total burn is eighty thousand dollars a day on salaries, office, insurance and so on. He says: if we can save one day, we have saved eighty thousand. If an employee asks whether it is worth spending twenty percent more so a part arrives five days earlier, he answers: absolutely yes, spend the money to increase speed. We will save more by going faster than by spending less and going slower. The link between speed and low cost is clear, the math is irrefutable. And further: if we go faster, we can run more experiments through fast iterations, we will reach quality faster and in the medium and long term get higher quality.

Hence the holy trinity: speed, quality and cost. These three things go together and do not oppose each other. The author stresses: going fast is not about working a large number of hours and not about burnout, but rather the opposite. It is about removing the mental and operational barriers on the path of speed. It is the opposite of working more. And the fact that you did not do this before is not your fault: you were taught the wrong thing your whole life and were not shown the right thing.

$80KStoke's burn per day 1 daya saved day equals $80K

Commandos, soldiers and police

The psychology of team behavior is broken down into three types of thinking. A commando takes the hill, finds resources, does whatever is needed. Soldiers set up barracks, run power lines along dirt roads, hang a pontoon bridge, calm the situation down. Police watch to make sure nothing goes wrong. If something goes wrong, that is bad. You have to pay a commando one hundred and fifty percent, and you will get ten times the return. Soldiers cost one hundred and ten percent and will be three times better. Police give less than one, they have negative productivity, they are very slow, but they still cost one hundred percent of the salary you set. Police should be avoided, and when you see any nearby, ask them to leave the building. A soldier should also be asked to leave.

For the first thousand people in the company, do not think you only hire commandos for the first five spots; you have to hire for the first thousand and want only commandos. A commando keeps a high speed bar, they have no fear, they simplify, they go where the movement is, they find resources: everything the whole episode talked about. If you bring in one soldier, he will bring in three. Steve Jobs said he would never hire a B player, because they bring in C players. It is the same with speed: hire a soldier and he will bring in police, because next to them he looks like a hero and a brave man. Hence the frequent question: I have eight good people and three slow ones, how do I speed them up? The answer: fire them, let them go. Thank God we are not in France, where firing is hard, and for that reason you should not set up a company in France.

Psychology and prioritization: what a founder really needs

In the early twenty-tens, when the author was between companies, several boards of directors called him to help with growth, because he was good at growing companies, and paid about 1.5% of the company for helping the CEO and the team with growth. These were Poshmark and similar companies. When he worked with the CEO, did he talk about growth? No, he talked about psychology and prioritization. A board of directors will never hire a person to help the CEO with prioritization and with his psychology, but it will pay for growth. And what is really needed? Psychology and prioritization. Founders of unicorns need help with their own psychology and with prioritization, and so do you. You have to find such people, spend time on them and set everything up properly. It is far faster to have the right priorities and to throw bad mental models out of your head, so that your people do not pick them up.

The prioritization tool is a priority dashboard, very simple. You hang it on the wall or project it, and change it in front of everyone: you write, others write. This keeps the team aligned, because everyone sees another's priority. Everyone has a pet project, and everyone gives presentations about why their project should move up higher. Every Monday the team gathers and asks: you are at the bottom in purple, justify why your project should rank above all the other projects. The answer "I cannot" means the person can go back to work and be faster, because they are no longer thinking about their pet project. It is very simple, takes about forty minutes a week and works great up to thirty people.

Customer feedback: the street and the mailing group

Do not wait for a focus group. Feedback can be had by lunchtime. The author's team carried a box out onto the street, showed the product on a laptop and simply stopped eight or ten people for half an hour to an hour, asked them to look at the product, give feedback, to understand what they clicked on and what they did not understand. Then the laptop went back to the office and in the second half of the day everything wrong was fixed. It cost nothing and happened instantly. All the introverts looked out the window at the author thinking "I cannot believe he talks to people." You can do this online, through a mailing group. That is how Gmail was done: the author was in the Gmail developers' mailing group, there were about eighty people, they commented and asked questions, and he was there for about fourteen months before launch. It is enough to simply start a mailing group, and people will help. You can use a Reddit subgroup.

The minimal experiment

You have to look for the minimal experiment. Most teammates will want to do something elaborate, because it is cool or because it is playing at business. You have to figure out what is minimal and get information without building the product.

The master of this was Bill Gates, the "king of vaporware." He announced products before they came out, and this kept venture capitalists from funding competitors, because it seemed to them that since Microsoft had announced it, it was not worth funding their competitor. So he had no competition, he simply announced things, never built them and kept his options open.

You can do everything on paper. The author's team constantly makes mocks, prints them on paper, puts them in a folder and sits down with a person to talk: what do you think, what would you do? You do not have to build. That is how Outdoorsy started: they were stuck, they were told to come up with a supply-side SaaS and talk to ten people. If eight, nine or ten say they want it, all good. If one or two, it will not work. If it is in the middle, we will run more experiments. They came back and said nine out of ten wanted it. Everything was on a sheet of paper. They built it, and Outdoorsy became a two-billion-dollar company.

Landing pages work too: do they want it, will they pay? You do not need to build, a landing page is enough. Do they click, do they enter a card number? You can even show a card processing error message. An explainer video works too: that is how it was with Dropbox, an explainer video about the product raised the waitlist from five thousand to seventy-five thousand people overnight. It was a three-minute product overview, but the product did not exist yet, there was only a good video, which maybe took three hours or even fifteen minutes, because the person just narrated over the mocks. Many companies started as email products or mailing lists. If you want to reach a customer base, start a mailing list. You can put up Google Forms.

The classic example is Zappos. They decided to sell shoes, took orders, got in the car, drove to a shoe store, bought the shoes, packed them and shipped them. That is how they did their first two thousand orders, without a warehouse and with almost nothing, through an ordinary form.

External deadlines and accelerators as an excuse

If it is hard to motivate people, create external deadlines for them. They will believe an external deadline more than they believe you, and will not want you to hurry them. Let the customer hurry them, the demo day, the release date. That is exactly why Y Combinator works: the deadline cannot be moved, everyone will come to hear your pitch, so you have to make it. At the author's accelerator people did twelve or eighteen months of work in three months. And the main reason this works is that they get an excuse in front of their entire network: sorry, I cannot talk to you, I cannot come to your son's christening, I cannot come to the birthday party, because I am in a three-month program and I have a demo day, I need to focus on the startup. And they become fast when they are not doing everything else. It is an excuse for the network to back off while you focus on the company.

Outsourcing, metric transparency and fewer processes

Outsource everything. Use AI, use outsourcing, stop doing what only looks like work. People have fear, they want processes so everything is done right, and this must be overcome. You have to explain culturally to everyone that there will be as few processes as possible, apart from trust and safety, payments and the like. And you have to make metrics transparent to everyone, otherwise people cannot act without you in the room. A daily email with all the metrics for the past day, sent to everyone on the team, is a top priority. Let everyone see what is happening.

What slows founders down: the long list

At the end of the module a list is given of what the author has seen slowing people down. This is a practical checklist.

A separate story about ambition and ego. The author talked with a fairly well-known person who wanted to raise five million dollars, and told him: the problem is that you want everything at once. You cannot focus for two or three years on one key customer, you are trying to do everything, because you have big ambitions and a big ego, and so I am not sure this is a good investment, because in my view you will ruin it all. That was a real conversation I had three hours earlier.

Module 2. Forty-two speed tactics

Episode 2 of 9 · encyclopedia of tactics

The second episode is a digest of all the tactics collected over fifteen years. Speed can be "built into" a startup as engineering, and if you learn and absorb these tactics, you will almost certainly manage to move fast. The material is divided into three categories: planning and thinking, team and culture, operations. The episode title says "forty tactics," but at the end the author calls them forty-two, because he adds two bonus tactical moves. Let us honestly note this divergence: forty main plus two bonus equals forty-two, and they are announced as "forty-two tactics you can implement today before lunch."

Planning and thinking tactics (1-6)

Team and culture tactics (7-18)

Operational tactics (19-40)

Two bonus tactics

After the forty main tactics come two bonus ones, and with them the author calls the whole series "forty-two tactics you can implement today before lunch."

First: make decisions fast. You have to distinguish "type one" and "type two" decisions. "Type one" are irreversible decisions, one-way doors, and you have to make them carefully. "Type two" are reversible decisions. Ninety-five percent of decisions are reversible. So decide, test and roll back if needed. Make a decision with seventy percent of the information. A large part of startup slowness is that leaders do not decide fast enough or aggressively enough. A startup is a psychologically violent environment, and with decisions and changes you have to be sharp: it feels cruel to you, but in time you will get used to it.

The second bonus tactic is the ethos of "say screw it more often." Across twenty years of running four venture-backed startups this changed the author's life. In moments when his back was against the wall and he had the courage to say "screw it," his brain switched to the right decision, he saw the correct path and could convey it clearly and forcefully to the team and the market. That is why his companies never failed: at some point "screw it" always rang out. It clears away old assumptions and plans for the future, a blank slate, and good ideas flood the head. Many wait until there are two or three months of capital left. If you learn to say "screw it" early, right after the money comes in, it will transform your life.

Module 3. Twenty-two mental models and the founder's personality

Episode 3 of 9 · psychology

The third episode is devoted to the second great lever after speed: thinking. After four startups and ten years in venture capital the author concludes that it all comes down to the psychology, mindset and personality of the founder. The four great levers of the course are language, speed, the right idea and mindset. Thinking is the greatest point of leverage: you can find a channel, you can find anything, but if you do not calibrate the brain, it will be hard. Contrary to the claims of some authors that anyone can be a founder, that is not so. Some founders are better than others precisely because of their mindsets, personalities and psychology. Here it is important to note honestly: in the spoken enumeration the author skips numbers eleven, thirteen and twenty, while saying that the list has twenty-one items, whereas the episode title promises twenty-two. We set out all the named models without forcing the numbering.

What venture firms look for

Large venture firms publicly state what founder characteristics they look for, because over twenty to forty years they have seen that these characteristics correlate best with success. NFX talks about speed. Sequoia about disagreeableness, that is, an unwillingness to be accommodating. Founders Fund about heretical and contrarian thinking. Hummingbird about neurodivergence. Yet another firm, which the author does not name, talks only about the candidate's relationship with his parents in the second interview, and it is one of the best seed-stage firms. The author separately explains why he understands psychology at all: his first startup was a psychological testing company, it had one hundred and fifty million registered users and twenty-five billion answered questions, with five PhD psychologists on staff. When he founded NFX, he began tracking founder attributes and built a framework of forty-eight attributes. Some are mindset, some are personality, some are psychology. The red ones in this framework are the ones discussed in more depth. Such a number of attributes shows that being a founder is not so simple: people who are truly good at this are fairly few, because to succeed in competitive conditions in large markets you need to possess many of these qualities.

Mindset versus personality

The course deliberately separates mindset and personality. A mindset can be learned: speed can be developed in yourself, and that was the first module. Personality is harder to change, and you should check your own. That is exactly why most of what is listed are models of thinking that can be mastered, not innate traits.

Models 1-9

Model 1. Speed

The first model is speed, and it has already been covered: speed can be learned. It is the foundation on which the other models are layered.

Model 2. Drive, including revenge

The question of the second model is: why are you doing this at all? Trying to prove to your father that he is wrong is a very good motivator. People try to prove they are worthy, that they cannot be ignored, that they are capable. Many are in the revenge game: a boss passed them over, a co-founder abandoned them, and they want to prove the opposite. Someone wants attention, someone wants love, because he loves working with a certain person. Someone wants not to disappoint a person he loves, a teammate. Someone wants to get into Wikipedia as a historical figure, to gain power, defeat enemies, earn money, give himself freedom. These are real motivations the author has seen in the CEOs he invested in. Elon Musk was already mentioned: Musk's friend (George Zachary) said that what helped Musk was that his father told him he would never achieve anything.

Model 3. The ability to manage a psychologically violent environment

Why is being a founder not for everyone? Because you can get post-traumatic stress disorder from it. The author says that after four companies he has real PTSD, and no one talks about this publicly. It is a violent environment, but the violence is psychological. Firing a co-founder, fighting a bad board member, losing a big deal to a competitor. All of this has happened to the author repeatedly, and to most founders too. You have to be ready to pass through this violence.

Model 4. Suffering

The author returned for the fifth Harvard Business School reunion and stood before the whole class. He made a confession in front of the whole class: he had just built and sold a company, had suffered like crazy, and no one had explained to him how much he would suffer psychologically, how little he would sleep, how many sedatives he would take to fall asleep and sleep three or four hours. He promised to help change the curriculum to warn people: if you choose the founder's path, a lot of suffering awaits you. That is the nature of things. You fail again and again, ninety-nine percent of people do not see what you see. You are alone, it seems to you that you are lying to yourself, you doubt yourself, you feel worse than another startup from the media and ask why not me. So much makes you suffer emotionally and psychologically, and that is the nature of the path.

Model 5. Commitment and burning the boats

"At the moment of commitment the universe conspires to help you." The same thought in other words: "there is power and magic in boldness." You must burn the boats, not give yourself a chance to retreat. Be fully in. Founders are often too good-looking, too smart and have too many options, and this hurts: they do not want to tie themselves down, because they are rational and have a choice. But for a founder this is wrong. If you remember the movie "300," the Spartans are told to flee, and Leonidas answers that they are committed and will stay and die, and calls it glory. That is the mindset of founders: total commitment at one hundred percent, not ninety-eight. The only way out is to go deeper. This is not the mindset instilled at school. School teaches you to keep options open: apply to six colleges, get four internship offers, date this person, but maybe that one too. You are masters of optionality, because you are rational, talented and well-off. If you want to do this, you have to stop thinking that way. Commitment raises the probability of success, and it can be learned: unlearn giving yourself too many options and go all the way, feeling the fire.

Model 6. Judgment, which is learned through many decisions

Judgment can be learned by making many decisions. In a startup there are far more decisions than at McKinsey or Goldman Sachs, so judgment develops faster. You can watch how others decide and learn from their mistakes and successes. Judgment raises the probability of success, and it can be learned.

Model 7. Tolerance for uncertainty

The literature shows that the two main predictors of durable entrepreneurship are a high tolerance for uncertainty and whether the parents were founders, that is, entrepreneurs. These two things were found in a Harvard study. Uncertainty and incomplete information are the only environment where a tiny startup, about to be crushed by an existing player, has a chance. So uncertainty should be welcomed. If someone on the team says "so what is the exact plan," the answer is this: there is the door, we are figuring it all out as we go. Uncertainty is needed, you have to enter it and love it, it is your only chance.

Model 8. Extreme ownership

There is a book called "Extreme Ownership," written by a group of Navy SEALs, and it is handed out to everyone at Stoke and some other high-performing portfolio companies. The main idea: no excuses, no complaints, remove blame and ego and fix it. If this happens with a person six times, he is fired, and before that he is not blamed. You have to ask yourself whether you are ready for extreme ownership or will blame someone else. "The partner did not do it" means you are to blame, because you did not foresee it. "The customer left, it is not our fault, we need more money" is wrong: the customer left because you did not anticipate the problem and ward it off and did not assemble a longer list of customers so that one departure would not matter. The author launched four companies, and all four succeeded, because he and his co-founders took extreme ownership of everything, and so there was no single point of failure and no failure.

Model 9. Be grounded in reality and ask for the negative

Real, direct, honest information is hard to get, and the only way to get it is to ask for negative feedback, to ask for a "no." The author was just at a board meeting of a company that raised money at a five-hundred-million valuation and wants to raise at a one-and-a-half-billion valuation. The whole board ran on "we are so great." But where is a proportionate degree of negative? You have to look at the company from two sides at once: "we kicked ass, our success came from doing it this way, so we will double down on it" and "we are in trouble, so we will do such and such differently and additionally such and such." You need both. You have to be grounded in reality, even if on the outside everything looks brilliant. Customers say what they think you want to hear. Friends too. When you ask a friend whether the idea is good, he will say yes, but not because he knows, but because he loves you and supports you. Employees want to support. Venture capitalists want to preserve optionality for future investments and do not want to offend you, and also do not want to earn a bad reputation. When the author first came to venture capital, he gave direct feedback, but that quickly ended, because people exploded and went crazy. About ten percent of founders "collapse" and run around telling everyone that "that guy is a jerk, do not work with him," and this gets in the way of attracting deals. People do not like interpersonal friction and do not give the truth. To get the truth, you have to ask for the negative, and the author strongly advises doing this in interviews.

Do not become Doug Burgum. He is the twelfth Republican candidate in the 2024 presidential election. He put in thirteen million of his own money, raised another four million and dropped out of the race, having spent sixteen point nine million. Why did he think he could become president? Because everyone around him said "I love you," and when he asked "do you think I should run," they answered "yes, I support you." He was surrounded by yes-men, he believed in electoral momentum and money. He was deceived, because everyone he talked to thought the idea was good. That is how his thirteen million went.

Models 10-18

Model 10. Strong opinions, loosely held

"Strong opinions, loosely held" is a way of talking with co-founders. In the nineties the author was told that a venture capitalist just needs to come to the board meeting, say "focus, focus, focus" and then leave. That is almost all a venture capitalist will say at a board meeting, because the founder's main problem is diffusion: "I could do this, I could do that, I have a customer asking for something else." So "focus, focus, focus" is the greatest value a venture capitalist can add to mindset, and even bad venture capitalists can do it, if they do not bury you in other nonsense. This mindset can be learned.

Model 12. Be contrarian

You have to be ready to be contrarian. The question: what truths do you acknowledge that few others yet acknowledge? This is Peter Thiel's whole line, independent thinking. It can be learned and it can be trained by reading people who think that way.

Model 13. The 10X rule

Many think: I will improve by four percent, by ten percent. It does not matter. Until you do something ten times better across the board, it will make no difference in the market when you are so small. If you are Microsoft and do something four percent better, everyone will say "cool" and adopt it. But for a small startup you need an order of magnitude, not percentages.

Model 14. Culture eats strategy

Drucker said: culture eats strategy for breakfast. Steve Jobs was obsessed with who sits next to whom. Jeff Wilke, Amazon's COO, said that the way we talk to each other is the core, and the language inside the company has to be carefully built. People in the know credit Wilke with building Amazon, because he walked into every meeting and every presentation and said: that is not how we talk about this, that is not how we do presentations, stop, here is how we think about it. He was very precise in language and in maintaining culture, and that is exactly what produced win after win. He became CEO after Jeff left.

Model 15. There is no corporate transformation without personal transformation

If you want a calm, easy life, take your Stanford degree and go work at McKinsey or Goldman. There you will be given tasks, you will do them well and live an interesting "normal" life. But if you want a real personal path, where you transform your mind, become the best you can be and welcome uncertainty, suffering and a violent environment, then this is for you, but you have to understand that you are engaged in personal transformation: running experiments and failing until finally one works. It is an experiment on yourself, and every change is a gift to yourself. If you go into a startup, you will grow in three or four years, not thirty. You literally triple your life by going through these paths.

Model 16. The risk of a mediocre life

You must fear a mediocre life more than failure. This should be your mentality. When you take a risk and fail, and this will be felt often, you must remember that it is normal, because at least you did not choose the mediocre path.

Model 17. Thirteen ways, eleven of which work

This came from Dennis Heightower and was said to the author and all his companies, and it changed his life: there are thirteen ways to do something, eleven of them will work, just pick one and go.

Model 18. Fundraising is about getting information, not money

There is a good article on NFX on this topic. Investors fixate on getting money, but that is too obvious. You have to get their information. Ask for a "no." If you get their information six, seven, eight times, then by the ninth meeting, if you ask both for a "no" and for their information, another venture capitalist you are now talking to will understand that you know everything and will himself say "may I invest." That is the process. Many fear "I have a great meeting, I do not want to ruin it." But you will ruin it if you do not get information. And remember that ninety percent of them will not invest in you anyway, even if everything is perfect.

Models 19-22

Model 19. Paranoia

"Only the paranoid survive." You have to stay paranoid, not become complacent and accommodating, not give yourself the right to success, not give in to fame, ego and self-congratulation. That is exactly what the company that went from five hundred million to one and a half billion was doing at its morning meeting. It is like rot and will kill the company if you do not help stop it.

Model 20. The commando mindset

This is about the commando with a high speed bar: simplify, go where the movement is, find resources. It is a separate mindset, already covered in the first module.

Model 21. The Goldilocks zone

Every attribute of a business has a scale, for example from arrogance to diffidence, and it happens that a certain confidence is appropriate right now, and that is the Goldilocks zone. But the Goldilocks zone does not mean the middle. If someone says "let us find a middle ground," for a startup that is nonsense. For example, some want naive founders, because they do not know what they do not know, while others say you need utmost clarity, sector expertise and knowing the customer inside out. But a person who has worked in the industry for twenty-five years often only does incremental things. So there is a balance between all these poles, and you have to seek the Goldilocks zone for each model separately.

Model 22. Toughness and determination

It is not about how hard you hit, but about how hard you get hit and keep moving forward. That is a quote from Rocky. Persistence, resilience, the ability to handle adversity. It is a mindset that can be learned.

Personality: the Big Five

Mindset is what can be learned, while personality is harder to change, and it is worth checking your own. The best personality test in the world is the five-factor model, the "Big Five," invented in 1972. No one has built a better test for segmenting and measuring people. The five factors are openness, conscientiousness, extroversion, agreeableness and neuroticism. There are thousands of human attributes that can be measured, but it is these five that separate people and define them most precisely.

Disagreeable people do not need everyone to like them and do not worry about what others think of them. This frees them for unique and heretical, unconventional thoughts, lets them pass more easily through conflicts and emotionally violent events and therefore more often get what they want, which means the startup too gets what it wants and becomes more successful. There is a big correlation between disagreeableness and success. Surprising correlations are also found with childhood trauma, neurodivergence, Asperger's syndrome and attention deficit disorder, and this explains why some firms look for exactly these traits.

Module 4. Magic words: language, positioning, brand

Episode 4 of 9 · language as a lever

The fourth episode is devoted to language, and it asserts something radical for engineering thinking. Everything in the world is done through words. The founder's job consists first of all in finding the magic words that lead to the product, to the customer and to the money. Language directly affects the speed at which the product, customers, employees and you yourself move. The founder's goal is to make something better, but you have to explain that it is better. It has to be what people want, not just what they need; and something investors will want to put money into. The mechanism here is this: explosively growing products with retention, that is, growth without churn, repeat revenue and customers, and this is achieved through language. Many will say: "What about technology?" Technology is needed too, but most miss that the game is actually played through language. It is almost as if the product came second and the engineering implementation third, because you implement what language gave life to. If you are able to make this mental shift, you will be far ahead.

First language, then the product, then engineering

Let us start with a textbook example. In a New Yorker cartoon a man stands with lemonade: "Lemonade, twenty cents a glass," and no one wants it. But "lemon detox" at twenty dollars a glass, and everyone lines up. It is a vivid reminder of how magic words change the difference. And it is not only in the consumer segment, it is also in B2B, and the rest of the episode consists of examples.

The main question to ask yourself is: what is this for the user? Where does it fit into his life? We conceptualize the world in our heads through language. What happens in your head is language: "what is he saying," "who is this guy," "where am I," "I am hungry." These little words run around in your head. So what it is for the user, that is the words, the magic words, that land in his head. People are busy: they have health problems, problems at work, romance, sports, caring for others, taxes, vacations, walking the dog. They are busy, and you live in a tiny crack somewhere in their day. You look at your product in admiration, while the customer sees it out of the corner of his eye, because he is busy with something else. If you are a VP of Sales, you have a lot going on, not just the software someone brings you. If you run an energy company, you have a lot of problems, not just this device. You have to understand what it is for them, so they will want to let it into their life, which is busy as it is.

Hence another famous thought: if you cannot explain it simply, you do not understand it well enough. When people come out with presentations and talk and talk, it seems to them that more words means more communication. That is not so. It is magic words, the right language, that communicate. You have to slow down: not more, but simpler. And it is not something you do on the side, it is what you do. Last time it was said that you run experiments; another view of the founder's job is that you are a language discoverer, a semiotician. All our reality is structured by language. The way our brains work is language first, function second. It seems crazy, especially when we are engineers, we love to build and bring something to the world, and we have in our heads that beautiful convertible everyone will look at. But language first, product second; language first, engineering third.

Language sits at the center of everything you do, because it touches your engineer, your hiring, who is drawn to you. People see words, read about you, watch you on YouTube, see how you speak in front of an audience. How do you raise money? What will you do on Saturday? Sales and marketing. Every part of the business touches language. It is at the center of what you do, and it is not something added at the end.

A whole company can be built on a phrase. How powerful is that? So powerful that you can build the entire business around the right phrase. The example "Find Old Friends" lived for about three, three and a half, four years and then faded, because everyone had already clicked the link. This also shows that language has a shelf life.

The power law in words

Another important thing about language: when you watch how the world works, a power law lies over the words used. If you use the word "free," whole books have been written about its effect on marketing and adoption. If you manage to use the word "free" or give the customer something free to try, that is a big deal. One cent is very, very far from free. Even one cent is a big distance.

An example from practice: a friend of the author ran a company called FitMob, which later merged into ClassPass. They tried thousands of phrases, and the biggest click came from the phrase "Weapons of Ass Reduction." Who would have thought. They grew for about eight months only because they could buy traffic on that phrase. Words like "sweepstakes" also work, and claiming that you are "number one." Being number one is the best way to differentiate from competitors. In your sector language also has a power law: if I use the word "database," I get one result, if "vector database," a different result, if "database of record" or "CRM," these are different words in your B2B space, but they have a power-law result in how people react and in what these words let you do for hiring, for investors, for customers and for employees. The main question again: what is it for the user and where does it fit into his life.

Reversing the order of product work

Most product managers, including the author early in his career, work like this: there is a problem, we come up with a product to solve it, then we market it, then we look for channels, where to go to buyers, and then we come up with the copy. That is, the copy is done at the very end, as if it were something small that can be handed to a summer intern. This is completely inverted. The right way is this: start with the language people want, put it in front of them, they will want it. Then figure out how to deliver it. "I will get you a date." "Oh, good." "Oh, how will I get him a date? That is hard, but I will do it, because he wants it." Then you come up with marketing and channels: go through Meta? At one B2B client of the author LinkedIn stopped working last year, and they are moving to TikTok and Instagram, seriously, to reach CFOs and VPs of Sales, because LinkedIn no longer works. Channels work for a while and then stop. Channels also have a technology window and fashion, everything changes all the time, and you have to align it all. And only once you have decided how to market can you build the product. This inverts how everyone was taught, because that is how it works more often.

Ringo: "store your photos" turns into "share your photos"

The company built a product at the dawn of digital photography, and the home page said "Store your photos," because there was nowhere to store all the shots taken with small cameras. Between 2,000 and 3,000 people came per day, but nothing interesting happened. The author thought: to become viral and grow without money, the product had to be multi-user, so people would share and market it themselves, not just use a free storage service at the company's expense. He went into the HTML and changed the text to "Share your photos." An hour later the team came into the office and said: "It says share photos, but our product does not do that." Three people on the team heard in reply: right, so go and make it. Now two or three thousand people come to you a day, and you are lying to them.

The team started fixing it. About a week later features appeared that allowed sharing and viewing photos. Over the next six months forty-seven million people registered for free, simply because one magic word changed. It absolutely changes your TAM, your growth and your customer. Besides, it changes the thinking about the product. If the product says "store your photos," and you hired a backend engineer with the ad "looking for a backend engineer to store photos," he arrives thinking you need to store photos, and when three days later you say "we are changing it to share photos," he is confused: "you lied to me." People get attached to words, turn on the news, and you can see how attached people are to words, it is one of the main things people do. So changing one word changes the thinking about the product, leads to new markets, to bigger or smaller TAMs and to new customers.

A community that votes on news: the missed Digg

There was a product the author described as "a community of people who raise news by voting," because he was thinking about network effects, about people and about community. Another person, about six months later, created a community-driven news voting site and called it Digg. Digg grew like crazy, then Reddit copied it, then Digg was undermined by its own community and killed, and now Reddit is worth tens of billions of dollars. They were so close. The difference was in the words and in the thinking. When the author thought about a community of people, he put the picture and the name of the person who posted the article at the center, because he considered the people in the community important. The whole team thought that way, and it was the wrong way to think, although the product was almost identical.

Help someone find a date

There was a product "find a date," but it could not attract enough people and could not buy traffic, because everyone said "find a date." Then a product manager said: let us make it viral, multi-user, and it is about helping people find a date. Now, if I am married or already dating someone, I can help someone else find a date, so I too can register and invite all my single friends. In fact you can invite all your friends, because even married or dating people can help someone find a date. In four months twenty-nine million people registered for free, and the business brought in about eight million dollars a year. All because the thinking changed.

The Dragons of Atlantis ad test

The company wanted to build a game. They went to Facebook and LinkedIn and bought ads. Some of the ads said "Wars in Egypt", "Wars in Xanadu", "Wars in Paris". The game's name was supposed to be "something something". After spending two thousand dollars, they discovered that Atlantis was the number one place where people wanted to play wars. They went off to make new ads: "The Amazons of Atlantis", "Dragons of Atlantis", "Realms of Atlantis", "Wars of Atlantis", to figure out exactly what it was. People said "Dragons". The author said: that is the game we are building, "Dragons of Atlantis". And the team learned exactly what it was building and what type of game. Within twelve months they reached one hundred twenty million dollars in revenue. That is how the magic words were found. At that moment a competitor came into the office and said he knew the company paid a quarter of what they paid for an install, but now they would launch a dragon game and start buying up all the dragon traffic. They had to grow as fast as possible, and in the end the company merged with those guys, because they were going to capture the magic phrase and make it more expensive. The business was eventually sold for eight hundred million dollars.

Snapchat, Yevvo, Air and Meerkat

There is a company everyone knows, and it had thirty four projects with many name changes. Evan Spiegel says they tried thirty four projects before Picaboo, launched Picaboo in July 2011, and in September renamed it Snapchat, and that is exactly when everything took off after thirty four attempts.

Another story. A company from Israel was called Yevvo, a terrible name: streaming live video to any other user. They worked on it for eighteen months and it did not work. The author said: if you move to San Francisco, I will work with you. They designed a new product, Air, a much better name: streaming live video only to your friends. It is a bit different, but it needed seven different features, and they did not finish building them, so the launch did not work. Then they came up with five different ideas, and Meerkat appeared, the best name of all: "broadcast live video to your Twitter followers in one click". The tagline already has the distribution channel baked in, because "Twitter followers" is a ready audience. First the language, then the channels, then the product. The technology, the backend, the people and the mission were almost identical across all these projects, but they had to find the magic word. Three hours after the Meerkat launch it was clear it would take off. The person had been obsessed with connecting people through video on mobile phones for ten years already, when it was hard, and the same mission only reached success after the words changed.

Positioning against competitors: hierarchy and status

There is a 1982 book, short, "Position Against Competition" by Jack Trout and Al Ries. The main thing worth taking from it is two ideas. The first is hierarchies and how the human mind works. Homo sapiens is hierarchical, we are tribal animals. If I am at the bottom of the pile, I will get less food and die, so I want to know my status, and it matters to me who is on top in my tribe. So a large part of the brain works on status all day long: half on optics, half on status. Once Coca-Cola established itself as number one and Pepsi as number two, it is emotionally hard for people to accept that the order could flip. So becoming number one in a category matters a great deal, and positioning yourself by calling yourself number one and reaching that position as fast as possible matters too.

There is a story about this. Hertz and Avis are the same as Coke and Pepsi. Avis's slogan was "We try harder, but it's number two". Because of this, and because the product was better, they eventually became number one. And what did they do? They bought a big ad in the Wall Street Journal and said "we are number one", they changed their television ad to "we are number one, we are so good that now we are number one". What happened? People did not like it emotionally, and they switched to number three. Then Avis came back and climbed back to number two, and everyone said: "Ah, I understand the world".

An apocryphal story is connected to this. Budweiser is number one, Miller is number two in American beer. Miller spent twenty five years trying to become number one. Then they came up with a new word: "light beer". They made light beer and became number one in light beer, went out and said "we are number one in light beer". People said: "Oh, this is not beer, this is light beer", and fully accepted it, because one word changed. Light beer was the number one beer for twenty nine years after launch, until Budweiser spent tens of billions on advertising, all those football commercials, to make Bud Light slightly bigger than Miller Lite. Instead of spending all that money to become number one, they created a new category with a new word. The exact same product, but they said it had fewer calories, and that was enough. It is the only product differentiation. The same color, the same bottle, the same factory, everything the same, one word changed, and they ruled for twenty nine years.

B2B, risk aversion and the first-mover advantage

In B2B the main problem is even more insidious than in the consumer segment. "You will not get fired for buying IBM", "you will not get fired for keeping Oracle". This continues because organizations are very risk averse and full of cops. Remember the cop persona from the first module: nothing bad should happen. So many B2B environments want to buy what everyone else buys so they do not get fired: "I did what anyone would have done, I did the sensible thing, nothing unusual". But there is an opposite effect here: once a company breaks ahead in B2B, it usually stays there. Both statements are true at the same time. Once you are number one, you attract more talent; once you are number one, you attract more capital for sales, marketing, big conference booths. This is called "preferred attachment", that is, a node in the network to which everyone preferentially attaches, whoever broke ahead. That is exactly why speed matters so much.

Verbs and buttons

Language sits at the foundation of growth. How do you grow? You spend money on ads, you hire salespeople, you get press, you go viral. But language sits under all of it, because ads and a salesperson's presentation contain language. If the language is right, the close rate is much higher, and the price you can ask is much higher. What will be written on the button: shop, find? The verb comes at the start of the sentence: "Find old friends". Which verb: find, list old friends, buy old friends? People really did try buying old friends with crypto: "buy your friend", "James costs twenty three dollars today", he went up, he went down. Any verb can be attached to any noun and turned into a business.

If the button says "share" instead of "invite", that is very different. If it says "add" instead of "follow", very different. If it says "friend" instead of "follow": one of those is Twitter, the other is Facebook. "Invite" versus "see who's in": one company's button said "Invite your friends", and another said "See who's in here!" with an exclamation mark. The first thought the second phrase was much better, tried it, and conversion rose by twenty two percent.

"Recommended" versus "people who bought this also bought". This is a story the author tells twice a week. In 1998 you are on Amazon, you see your product, and at the bottom it says "Items we recommend for you", there are three products, and none of them are needed. You think: these guys are just trying to push more junk on me, they are not helping me, they are not on my side, and they are dumb, because I do not need any of this, I really do not like them. But I will buy anyway, because I am already here, only I will not come back again. Some time later they changed the phrase to "People who bought this item also bought this item". And after that: people are dumb after all, I would not buy these things, but never mind, I do not care. They were not recommending to me, they got themselves out of the way and just gave data. The language change completely transformed the Amazon experience.

Product language needs emotional hooks. "Create a Profile" is worse than "Start the Action!". Everyone says "create a profile", but you can say "start the action". Explore the studio space, explore the language, look for emotional hooks. And remember: it has to be about the user. The company Honk: "Uber is an app for people who do not have a car. Honk is an app for people who do have a car". Immediately clear: this is about me, I have a car, maybe I need this Honk app, whatever it is. Simple and clear. Make it about them. "The language has to be about the user, not about you".

The author works on this for hours with founders. "Signing in with Facebook will make our app work better for you" versus "Feeling lazy? Sign in with Facebook". The second is about me: I am lazy, I will sign in with Facebook. The response: "James, aren't you telling them about the benefit for them?" You pretend it is about them, but really it is about your ego. "Feeling lazy? Sign in with Facebook" is much better, and when they switched to it, conversion rose a lot.

Slack versus HipChat, Anduril, Simon Sinek

Under the hood Slack looks like any other chat app. HipChat existed six or eight years before Slack launched, absolutely the same product. But they designed it differently and changed the language. The language sounded different: "Slack behaves like your witty robot assistant, instead of the boring corporate chat it could have been". It was given the color scheme of a video game rather than a corporate collaboration product. Everything seems to bounce playfully; the interactions are designed not only to help you understand what is happening but also to make you smile. That was the only difference. The same technology, the same engineering, the same B2B product, deeply embedded in Fortune 500 companies today, serious stuff. We are tribal, hierarchical animals, and we tend to relate to everything as a person. So you have to think about the personality of the product. That matters to people, and product and engineering people often do not think about it.

Can it be copied? Too late. Everyone has already chosen their robot assistant. Slack stole the show, everyone copied afterward and could not catch up.

Anduril uses the words "software-first, speed, venture-backed, AI, autonomous, Silicon Valley" and the phrase "Rebooting the arsenal of democracy". That is not typical Raytheon language. The company was named after something from "The Lord of the Rings", full geekery. They stand against the traditional defense contractors. They call themselves a "merchant supplier" against "prime contractors". They thought very carefully about the words, and they sell missiles, and it matters. It is the hottest defense company in the world right now.

There is a well-known thing from Simon Sinek about "why". He says a company should have a "why", then a "how", then a "what". Most sell the company from the "what" (what they built), then the "how", and only maybe the "why". But Apple positioned itself by saying: "we believe in breaking the status quo, we believe in thinking different", and they do not even use the adverb, because they want to say "different", not "differently", just as they want to think differently. Because it shows up in everything they do, that is why Apple is Apple, while other companies only do "what, what, what".

Who is responsible for the words

Who is responsible for this? Can it be hired out? Can it be handed to an agency? The answer is no, because only the founders can change the words on the home page. Only you can make the salespeople use different language in presentations. Only you can talk to the journalists who will write about your company. So it is on you. You have to be obsessed with language. It affects every facet of the business. It is your product. If you are not strong with language, find someone on the team who is strong, and lean on him right now, give him co-founder status, whatever it takes. Someone has to be sensitive to language and worry about how it feels. You need a process with spreadsheets and A/B tests, talk directly with customers, try to feel how it sounds. Set up a process and pay attention to it.

Find the right word, like "share your photos", and you immediately have forty seven million users. If you do not spend the time finding the word "share", you will grind for two, three, four, five years and wonder where your life went. "Where did my twenty years go" because you did not find the right word. There is another way to look at "fits". The founder story, the company name and logo, the words and language you use, this is "product founder fit", then "product market fit", and then users. Words touch every part of this. You are part of the story. You are part of the language. The question is: what can you convincingly lead?

About copying: language has a shelf life, like "Find Old Friends", but so do the things everyone copied. Google invented nothing, copied everything, and even AdSense was bought, not created. Facebook copied Friendster, Airbnb is VRBO done differently, Uber copied Lyft. So borrowing is normal. But language is still unique, because you and your situation are unique.

Module 5. Speed on the market map

Episode 5 of 9 · terrain reconnaissance

The fifth episode is devoted to the market map. Most founders underinvest in developing the market map and in understanding the terrain they operate on. Many investors say that if they had to choose between founders and the market, they would choose the market first, because even mediocre founders can perform well in good markets. This may be true for some investors, but it shows the power of a well-chosen market. Understanding your terrain and mapping it correctly lets you move through it much faster. The biggest mistake founders make is not understanding the competition. You cannot drive fast if you do not know the terrain you are driving on. What kind of car do you need for this terrain? Where is the decision point to turn right or left, a strategic move or a product move? Without knowing the terrain you will drive slowly, you will not have the right vehicle, the right moves or the right knowledge.

Why the map is needed and why you will not drive fast across unfamiliar terrain

The goal is to become the smartest person in the room. You have to be able to make fast decisions accurately. You have to know the landscape. You have to lead the whole team and everyone outside the team. So you have to know more about current and past competition than anyone else, because you do not want to repeat their mistakes and you want to borrow their best ideas. They are literally your outsourced research and development department. Knowing all of this helps you attract teammates, customers, investors and press. In a sense you have to be Indiana Jones: you have to study and be the smartest, to deeply understand the history of your sector. You are the guide who will take all the employees and all the investors through the jungle, and you have to give everyone else confidence.

It is known that investors spend an average of three minutes and forty four seconds reviewing your deck when you send it. And what do they spend the most time on? The data shows that almost two minutes of that three minutes forty four seconds they spend looking at the competition map, at the competitors slide. Why? Because they need to understand your company's positioning. Maybe they see that you are in a dead end. Maybe you are racing toward a brick wall. Maybe they already know about eight other similar companies that failed and why, and it simply will not interest them if they know a lot of this. Or they will be able to see how it fits with other companies they know. So the competition slide you give investors is the most valuable source of information for them, which means it should be the most valuable for you too. And you have to construct it.

It is 2026-2027 now, this is not 1994. In 1994 the browser came out, and you could get away with not knowing the history and the competitors, because there were none. But the times are not the same. This is not the time of the movie "The Social Network". You have to know what came before you. Most of the ideas you are going to try have already been tried by some team in the recent past, and investors have seen many such pitches. Some think: "I do not want to get fixated on competition, I want my own vision". Keep your vision, but do not be naive and do not be ignorant about the competition. You have to know it by heart, because it will help you move through the terrain faster, aggregate resources to the company faster and move through the landscape.

Another mistake: saying "these guys only do X", "they are not very smart, they only do Y". But they are not idiots. They think they cover the whole space, just as you think you cover the whole space. Your competitors use the same language as you and thereby signal to the market that they cover the whole space. Their customers may think they are everything, not just the little square you placed them in. Do not do that. Do not say "they only do this". They almost certainly do more than this, they have good investors, they are not stupid and they understand things.

The four parts of the market map

The market map has four parts. First: the competitor table, which you review regularly, every week of the company's life. Second: "two by two" chart matrices, and in the plural, not just one. Third: other charts for yourself, for employees, for prospective employees, for journalists and for investors. Fourth: understanding who your customer really is, which leads to estimating the serviceable addressable market (SAM) and the total addressable market (TAM).

Where to gather data

Tactically, data comes from ChatGPT, Claude, Gemini, Google. From Crunchbase, DealBook, CB Insights. These services have already built databases about your competition and its history. Read industry articles. Talk to investors: when you meet with investors, be sure to spend part of the time asking what they have seen in this space, which other companies yours reminds them of. They will give you a pile of interesting data you did not know, and you have to record it in the table and in the competitor knowledge base. Talk to customers and ask them about the same thing. And, importantly, talk to competitors. One of the main things the author taught people: you have to talk not only to customers but also to competitors, or to those who used to work at competitors. Look for them on LinkedIn, these are former employees. Invite them to lunch, to dinner, buy 49ers game tickets, whatever, and just ask questions: what did you try, what worked, what did not, what was the history of this. You will learn a great deal about the probabilities of what may happen to you further out on the terrain, and that is exactly what will make you fast at making decisions when new data arrives. Go to industry trade shows. And the thing people constantly forget: archive.org, the "Wayback Machine", go in and see what language competitors used, what their onboarding was like, how they positioned and pitched themselves. This fills in the market map and makes you the smartest one in the room.

What goes into the competitor table

An example of a competition map is far more detailed than people usually think. You have to know the company names, how they changed names, who the investors were, where they are located, how many employees they have. It has to be very detailed, it is worth using colors to break it up, but the level of detail has to be almost extreme, you have to know everything.

Separately, about "what they do well". You have to figure out what your competition does well, instead of burying your head in the sand and saying they are stupid. They are not stupid, they are just like you. They do something well. Why do they have success, why did they raise the money they raised? Because they do something well. Admit it, understand it, maybe borrow it, learn, benefit, use them as an extended research and development department. Do not pretend they do everything badly and that you will destroy them, just to have the courage to get up every day. Understand this mental model and accept an understanding of the competition. The level of detail in the tables has to be extreme, and you have to keep it in your head so that in any conversation everyone sees that you are the smartest in the room. Interestingly, when you build the market map, writing it out and spending hours on it, it literally goes into your head, and that makes you fast.

The substitutes tab

The table has a second tab, the substitutes tab. These are not direct competitors, but you have to admit that butter and margarine are not direct competitors, they are substitutes. Or entertainment: watching YouTube is a substitute for playing games. They are not direct competitors, but substitutes from the point of view of how people spend time. Solar energy versus the energy of coal plants. You need a tab where you fill in the companies that are substitutes for what you do, to understand their business models, whether they are better or worse than yours and how they approach the market. When you speak more broadly, with journalists or investors, you need to understand these substitutes, not just direct competitors.

"Two by two" matrices and other charts

The second thing to build is a series of "two by two" chart metrics, to visually understand where you are. Usually you always want to put yourself in the upper right corner. But there are many ways to slice markets, so you have to constantly iterate on the axes. The author has seen companies with six such matrices, to look from every angle and identify who their competitors are, who the substitutes are, who everyone is in this landscape. It helps founders keep this in their head, explain it quickly to any listener, and literally walk up to the diagram and point: here we are, here they are. Such communication is fast, there is no argument or hand waving, everyone calms down and understands what is what. Remember, investors spend most of their time on the competition slide, and there is a reason for that.

The third thing is other types of charts and visualizations, so that you, employees, investors and journalists understand where the company is in the overall landscape. For example, a horizontal bar chart, or the charts that CB Insights makes, they have amazing charts. When such charts are up on the wall and show where you are, it is easier for people to locate you, because each listener may be talking to a robotics company, a B2B software company and a space company, and they need to understand exactly where you are on the map.

Customer, SAM and TAM from the bottom up

The fourth thing is to know customers well. You have to be able to point to your core segment, to the secondary segment, and to those you are definitely not going after in the first or second phase of the company. This helps you be clearer and make decisions faster, because you understand what you do and what you do not do. You have to know whether the market is big enough, and see whether you are growing into it fast enough.

How do you build a target market? The initial target market is a small group of people you start with in beta, in alpha or in the first months. The serviceable addressable market is the next layer, and taking competition into account it shows what you can really go after. And only once you have beaten the competition can you go after the total addressable market, that is the terminal value you can pursue, and it lets investors and employees understand how big the opportunity is. It is the serviceable part that requires accounting for competitors, because you will have to beat them to get it.

When calculating, do not make the obvious top-down mistake: "the market is two hundred billion, if we take one percent, we have a two billion opportunity". That makes no sense to anyone. You need a bottom-up analysis. How many salespeople can I have? How many sales can they make per week? A detailed bottom-up analysis, not "the world is big, pie in the sky". This is part of building trust and leading people: how we will build this into the company everyone imagines.

Module 6. Speed to revenue: pricing and conversion

Episode 6 of 9 · money and behavioral economics

The sixth episode is about how, once you have revenue, everyone will look at how fast your revenue grows. That is what will define your speed in the eyes of those around you. So you have to work through the topic completely, quickly and point by point: why you sell, why people think they are buying from you, whom you sell to, how you sell and when you sell.

Not all revenue is equal

Realize a simple idea: not all revenue is equally good. Some revenue is better and some is worse. When the margin is high, such revenue is better. When it is recurring, better. When it grows fast, better, it is worth more. And if your revenue appears to be protected by patents, network effects or being embedded in the customer's processes, it is worth more. The right path will ultimately be simple, though at first it looks complicated. At the start there is a lot of complexity: many ways to take money, many ways to set a price, many ways to present it, many channels. But in the end it all comes down to something simple, and you have to pass through that complexity.

The "why people buy" hierarchy

One of the things the firm studied for a long time and has not yet published is that "make me money" is the best way to sell anything. If you give me twenty dollars, I will make eighty on it this week, why would I say no? Another example: I pay you 2,400 dollars, and you make me 7,200. You make me more money. Then comes the hierarchy of motives.

Many build businesses that help customers save money. But if you come in and say "I will save you money", you get the answer "maybe". And if you come in and say "I will give you a device to inflate tires better, and you will earn an extra two thousand four hundred dollars", the person sees it as revenue and says "yes, I want this". That is clearer. "Save money" is perceived by everyone as "maybe", I am not rewarded for this, no thank you. And "save time" is valued by almost no one. The hierarchy of motives helps you understand how you set a price, how you monetize, when you sell: everything else flows from understanding this. It applies to consumer products too, to almost all of them, except perhaps the state, but even the state usually follows this.

Whom to sell to: the side that pays, and the order of targets

There are many people who can pay, but you have to know the types. The marketplace side that is easier to attract is usually also the side that will pay. The consumer pays for himself. The state pays to protect its citizens. There are indirect sales. For example, in healthcare it is often the insurance company that pays, not the doctor and not the patient. The one who consumes does not pay; the one who administers does not pay either; someone third pays. A very strange, very unusual market.

An example of indirect payment is the company Zenefits, which for some time was the fastest-growing SaaS company in history, because it gave its software away for free. Who paid? Health insurance companies, they paid thirty dollars a month for each person using this HR software. The money was taken from the insurers, who paid commissions on all the health insurance sold through Zenefits. Companies said: "I get this for free, I get this HR software for free and I have to pay for it?" They were answered: "yes, because in fact we are an insurance broker and earn indirectly, through a third party". You have to think: is this possible in your industry, can you get someone else to pay.

A separate example: a bank pays for something to be installed, the company gets it for free, and is even financed by the bank. The direct user is not always the one who pays. The cost center does not necessarily pay: maybe his boss, maybe the CFO, maybe someone who holds the budget, not the CFO but the budget owner. And there is an order of whom to sell to: if you can, sell to the CEO. Then to the VP of Sales. Then to the CFO. And at the very bottom, HR, the worst place for a B2B sale. You just have to be aware of whom you are selling to, and explore this space.

How to take money: subscriptions, usage fees, the willingness-to-pay curve

Let us start with one type of charge: subscriptions and usage fees. Picture the customer's willingness to pay and the number of customers. If you charge a subscription fee, you are setting a price. The people who do not get this value and are not willing to pay that much stop being your customers. Only those whose willingness to pay is above the price pay, and they pay a fixed amount. This area is your total revenue. And this is pretty bad, because these people could get some value if there were a way to charge them less, and over here they extract a lot of value from you, and you do not get paid for it. Bad.

So you add usage fees to reach the "whales". This can be per seat, per API call, for a superpower, for storage volume (you want to store more, pay more; you want more storage, pay more), per transaction, for revenue processed, per location, for acceleration. Acceleration is when they say "it will take us three months to get to you, but if you pay more, we will do it in a week". In some countries this is called "graft", acceleration. These are all types of usage fees, there are others too, but this gives an idea. Then usage fees start to approach how much all these people are willing to pay. Now your revenue is this entire area, because everyone pays according to their desire and their willingness to pay, with a usage fee, if you design it smartly. Theoretically you can create cheaper subscription tiers and maybe capture a small piece over here. You have to try to extract as much of the value you create as possible, through the price, through how you take it, from whom you take it, when you take it.

Gross margin for real

Gross margin is the price minus the variable or marginal costs of delivery, and this includes all variable costs, not just the obvious ones. Most people do not add in all the variable costs, including raw materials. You had to build tire inflators, deliver them, and delivery of these heavy metal things costs about two hundred twenty dollars, and that comes out of the profit. Salespeople, that case with 250,000 dollars that was mentioned. Sales. Marketing: you had to buy a booth in Vegas at the tire show, buy ads on Facebook. Financing costs: you have to pay interest. Variable labor, when something goes wrong. All of these costs are variable, marginal costs of delivering the product or service, and they have to be included in full to understand the true gross margin, because it is out of it that salaries are paid, it is out of it that the profit comes, which gives your shares their value.

If you calculated the gross margin and ended up somewhere around forty percent, even though you thought you were a software business, then you are not a software business and you will not be valued as a software one.

Pricing methods

The first and simplest way is to look at existing price benchmarks. When you start a business, price a similar product the same as the existing benchmark, whether it is a direct competitor or a substitute. More expensive, and you will not understand why it does not work. Cheaper, and you lower the perceived value and waste vital margin, especially at the start when you need to hire one more engineer. Lower the price only if your costs can also be lower. The whole business has to be built around the pricing strategy if that strategy is critical, and this is a holistic strategy. Do not go down out of fear. In fact the last slide of the presentation is "raise prices". You want more revenue with more margin. Build more value to justify an equal or higher price. Most businesses worth building have higher prices. Do not enter a business where the only way to win is to grovel around your own price. Find a business where you can charge more and everyone will applaud. That is exactly how big businesses are made.

The second way is to orient toward costs. You need enough gross margin in the product for the business to thrive, so the minimum price is calculated from the costs. The third way is value-based pricing: the price is a share of the economic value you create. Often this is five to twenty percent, typically ten percent. Sometimes the author says: "I will save you five hundred dollars, I will take one hundred from you", and the answer is "yes, fine".

The fourth way, and a very interesting one, is the willingness-to-pay interview, which you can try right now. Interview four to ten potential customers, show them a paper prototype, a Figma, a PowerPoint or a picture, describe the benefits and ask at what price they would feel: one, that it is suspiciously cheap; two, that it is a good deal, no questions; three, that it is expensive; four, that it is prohibitively expensive. Four questions, at what price each of those four feelings would occur. Then plot the answers on a chart, and you will see the point that maximizes revenue. Then validate that point with fake landing pages and checkouts. It looks like this: too cheap, cheap without question for the value, expensive, too expensive when they say "go away". And you will see the price-maximizing point. This came from the book "Monetizing Innovation".

The fifth way is tiering, different prices and levels, and observing what people do. You can just put up a page with tiers and watch what happens. Then you realize something is very bad for you and remove it, or you discover that no one buys something and you adjust. This is a way to learn the price through tiers.

The sixth way is channel-based pricing. If you need resellers, retail, brokers, app stores, partners, sales reps, the price has to leave room for their margin, for their customer acquisition payback, for their commissions. If you use salespeople, you have to leave room for them to earn too.

Behavioral economics and the presentation of price

Now the most interesting part. If you anchor at 142 and say that for you it is 92, conversion rises. If you use nines and discounts, this is called charm pricing, or the "left digit bias". If you charge 39 dollars, you have a certain conversion, and it will rise, because you anchored and said it was a sale, and a sale may be urgency: it will not always be on sale. So conversion can be improved noticeably while taking the same amount. But this may not work for your brand. Nordstrom does not do this, because when you come to Nordstrom, you expect to pay: a blazer for one thousand seven hundred dollars, no 16.99, just one thousand seven hundred, 2,400. You came expecting to pay, so there are no discounts there. For discounts, go to Ross or Nordstrom Rack. Be careful: discounts may not suit your brand.

Another technique is the "decoy", the third choice. Here is The Economist: a magazine subscription for 59 dollars, or print plus web subscription for 125. Fifty nine or one hundred twenty five. Sixty eight percent chose the first option, not brilliant. Then they changed it: they added a decoy, the print subscription, and then print plus web, and eighty four percent paid 125 dollars a year just because a decoy appeared, and of course no one chose it, because it is obviously stupid.

Another example: if you sell two sizes of popcorn at the cinema, large and small, or large, medium and small, you will get seventy percent more revenue simply because of the decoy. With wine it is slightly different: two bottles, at 10 and at 30, but then they add a decoy on top, and the 30 seems reasonable. This is the concept of "second largest or second from the bottom". On a wine list people almost always take the second from the top or the second from the bottom: I do not want to be cheap, but I also do not want to be stupid, I will take something good that will suit everyone, depending on who I am with.

Another technique is to test the boundaries. The author ran a gaming company, and there you could buy rubies. There were subscriptions, but money was taken for rubies, so you could attack more, defend more, play smarter. They sold rubies, and that was the main revenue. They kept this tiering for about a month and a half, and the author said: to hell with it, set it at 499. The team replied that this was madness, wrong, abnormal. They set it, and within an hour someone bought. Then about forty percent of revenue started coming from buying that tier. People paid ten thousand dollars a month to play the game. You have to push the boundaries, learn, take more, see what happens.

Social proof: "most popular", you lead me with social proof. This is one way. There are also overcomplicated pages that serve as an example of several things at once: yes, you should experiment and see what happens, and then compress it to something simple. There are very ugly overloaded pages, although the social proof on them is good and positive, the overcomplication gets in the way. Simplicity usually works. Google had its own page when they first launched payments and no one paid Google for it. It was their first way to take money: simple, not 4.99, we are Google, we are above that, simple and calm, no serifs, very good.

Have you heard of choice paralysis? There is a book, "The Paradox of Choice", and it analyzes an experiment. In a grocery store they set up a table and a person selling a new kind of jam. There were six kinds of jam: with six, forty percent of passersby stopped to look and thirty percent bought something. Then they put out twenty four kinds, a whole cacophony of jams, everyone came over to look, sixty percent stopped, but only three percent bought, because their brain "locked up": too many options, I do not want to choose wrong, I do not know. Usually three to four is optimal.

Perceived value and price are value. The more expensive something is, the higher the perceived quality and the higher the actual efficacy. People were given a placebo for 2.50 and Prozac for 2, and the reported efficacy was higher for the expensive one. Another example: a person comes up to you with a cooler on the beach: "want a beer?", "yes", "I will give it for three dollars", "fine". Or you can walk two hundred feet to a shack with leaves on the roof, and there: "I will give you a beer for fifteen dollars". And you are like: yes, I am sitting in a bar, I look cool, everyone around is dressed up, the television and lights are on, fifteen dollars is fine. The same beer, a different context. How do you create such a context for the customer that he is ready to pay fifteen instead of three? Expensive fashion stores cluster together, because then the thinking arises: "that scarf is 2,500, this bag is 4,500, but a belt is only eight hundred, that is a bargain", even though the belt costs them four dollars to make.

Social proof raises conversion: "more people in more places manage more projects online with Basecamp than with any other app", "last week this many companies launched new projects". This creates conversions in the interface, in PowerPoint and at the meeting: other people are doing it, you will be safe, you will not get fired for this. Give numbers. Show the faces and names of friends I know. If the site showed the face of James Currier, whom you hang out with at Stanford, and an exact quote, you would believe it, because you know that person and his name. That is very different. Now software can do this: people I admire, people like me, experts, many people. This worked for others before, do it again.

Another technique is framing. "Only 10 percent fat" versus "90 percent fat free". How much ice cream will you sell? You want to sell the second. Go positive, name the benefit. Another framing: if you give people these options, eighty percent more chose the third. If you put this in an A/B test, eighty percent more people will click on the free product. "Free product, only five dollars for shipping": wait, wait.

Scarcity: limited quantities, limited time, rare and ultra-rare. This is a phrase the author's wife and children now repeat: take anything and make part of it rare, part of it ultra-rare, and simply take more. Streetwear, sneakers and the like. The king of all these techniques is booking.com, they use every trick of behavioral economics to make you buy at a high price. People ask how to learn more, and the author answers: go play on booking.com. A good source for going deeper is "Predictably Irrational" by Dan Ariely and "Influence" by Cialdini. These are classics, the author rereads them every two or three years.

The power of free

The word "free" releases large amounts of dopamine, and we react very irrationally. "Buy one, get one free" gives thirty to fifty percent higher conversion than a fifty percent discount, precisely because of the word "free". Why are there trial periods and customer pilots? Because free is plus twenty percent. The only difference between one and the other is that it says "free". Plus twenty percent, plus twenty eight percent: 18.6 versus 14.5 percent conversion. Free moves the needle dramatically.

There is an experiment about fourteen cents. One option is fourteen cents more expensive than the other, but if you put them on a table in a supermarket and ask, seventy three percent will buy the expensive one and twenty seven percent will take the cheap one. And then you move to free, and they still differ by fourteen cents, and the picture flips. However, you have to be careful with free: you can attract cheap customers, and you do not want that. They come for the free and do not convert to paying, they can spoil the customer relationship and the brand. Watch the metrics at the bottom of the funnel.

On the topic of free there is a whole book, "Freemium" by Dave Boyce, who was at many very successful companies. Giving a free trial period or freemium is better than just free. You should not give things away for free, you will attract cheap customers, but you can give a free trial period.

When to take money

When do you give a free trial period, when do you stop it and how do you justify it? You have to look for a natural breaking point, and many businesses fail because they do not have a natural point of transition from freemium or a free trial to paid. Experience does not give a clear break, but you are looking for a clear point when it feels natural. Good examples: your storage has reached its limit (Dropbox), perfect and simple. Add one more seat per person. Your thirty-day free access ends in seven days: you write a letter "seven days left, pay". Get it without ads, stop the ads, YouTube does this now. You pay zero, but ten percent per transaction: Airbnb eventually became a copy of VRBO, but changed the model from "pay three hundred dollars a month for a listing" to "share ten percent of revenue with us". That was the big change, the natural point of transition from free, when a transaction appears. "We pay you and keep the profit, we will buy your house, but on resale the profit is ours". Maybe you take no money from the first hundred people at all, because you want them on the site and using it, and from the 101st you start charging right away.

Other examples: pay for promotion. On a dating site they say: if you want to be at the top of the list when women search for men, pay. Listing is free, but if you want to rank and be found, pay. 99Designs does the same. Waiting-time acceleration: any video game essentially charges you to go faster: the factory produces a certain amount of coal per day, you do not want to wait all day, pay for acceleration. Security or enterprise level: want more security, pay; want integration with the IT department, pay. Tips: there is a company EarnIn, I think around seven hundred million in revenue, it is a payday loan business, you withdraw money for free, but "just give us a tip for doing this for you", and about seventy percent of people leave a tip, and that is where they make their money, having given the service away for free. Three-degrees emails: LinkedIn does this, you use the whole product, you can message first and second degree, but not third, pay three dollars a year. WhatsApp: pay to keep using it, free until the network effect kicks in. If you do "free until network effect", like WhatsApp, you have to be ready to raise a lot of money. It is an organic part of how you design the business.

There are companies whose products are so good that they do not need free trial periods. They charge for the whole year up front. After a fourteen-day trial they take an annual subscription, to finance the company's growth, they need less venture capital and give up fewer shares. They take the full year up front, like the company F2 in New York, about a million dollars per customer.

The final lesson: raise prices and find the strange willingness to pay

The last slide again: raise the price. Do not keep prices low out of fear. Become a monopolist, if you can. Find a product people will pay anything for, work on a project like that, not on the hard ones. Look for the strange willingness to pay. People have strange attitudes to storage: they will keep their junk in a warehouse forever, eighty five dollars a month, two hundred a month, three or four years, just not to lose things. We use this strangeness: in games, if you want to store your shields, spears and guns, if you want to store them for the next week, you have to pay. People are like: yes, I do not want to lose, I do not want to lose my things. That is strange. That is why it was so easy for Dropbox to take money. Coffee: a cup of coffee cost 86 cents, and then Starbucks comes along and says: no, no, five dollars, and you are like: fine, five dollars, I will pay anything for that little boost. Now it is seven dollars, and nine if you want pumpkin. And then poker. Texas HoldEm Poker was the foundation for Zynga, which came to be worth eleven billion dollars. It is where you put in real dollars to get chips, but the chips cannot be cashed out. It was not even a poker game, because you could not win money, which is why it was legal. People played poker and simply accumulated chips for status. It is madness, and on this strange mental leap that "fine, I will just pay more, I want more chips", Zynga built its whole company. And take prepayment if you can. And if you make people more money, it is easier to charge.

Module 7. Fundraising speed: the ladder of proof

Episode 7 of 9 · capital and network

The seventh episode teaches you to raise money with maximum speed and minimum effort. Many founders say they just want to get through fundraising and get back to building. That is legitimate, but the fact is that a founder is fundraising all the time. To move the whole enterprise forward, you have to be able to fundraise with the least effort and maximum speed.

The ladder of proof

The key concept of the module is the ladder of proof. Think of your business as a ladder of proof. The investor asks himself: were you able to attract a great team? He will look at your LinkedIn, get a referral from a mutual acquaintance who says these people are excellent. Then he will ask: and what are they doing? He is told: here is the idea. "Oh, they are doing software as a service, that is timely and hot, fine, maybe I will meet". Then he will ask at the meeting: have you identified the fastest-growing market, or a fast-growing market, or a big market, how does he see it? Then he will ask: who is helping you, where does your expertise in this area come from, has the team proven it can build a product? All of these are rungs of the ladder. Maybe you have not yet proven you can build a product, but you have expert advisors and other things, and that is where you are on the ladder of proof. And it continues: do you have customers using the product? Are they staying? Is someone paying? Is there revenue? These are all rungs of the ladder of proof. In all conversations with investors, journalists, employees you are asked exactly about this, and the question is essentially: where are you on the ladder of proof and does that match my risk profile. All the way up to: do you have rapid growth.

Rapid revenue growth is a unique rung, because once you reach it, all the other rungs seem to fall away. If you are growing very fast, you probably have a good enough team for that to happen, you have already built a product, you already have customers, they are retained, you have found a scalable channel. All these questions are answered by rapid growth. That is why there is so much focus on it.

Interestingly, on every rung where you are, there will be an investor who wants to invest exactly on that rung. Where you are is not wrong. It may be wrong for a particular investor you are talking to: he will be disappointed when you say there is no scalable channel yet, and he will say he cannot invest then. Fine, you are looking for companies at higher rungs. And conversely, there are those who look for low rungs: "Oh my, you already have revenue? We never invest in people with revenue, that is amazing, you have come so far, I am thrilled". That is how they see the world. So your task is to find investors who match where you are on the ladder of proof. This realization came once when the author was asked whether the company already had a name. He answered that they had not yet named the company, and he was told: go away, come back when you have a name. He was constantly being turned away to make him climb one rung higher. On signal.nfx.com investors are listed by stage: pre-seed, seed, series A, series B, as well as popular lists, for example, you are a woman, you are in Israel, you are from LATAM. All these investors raised their hands and declared which founders they love to fund: this geography, these types of people. Finding the right investor and matching your place on the ladder is exactly how you raise money at each stage. You should not be ashamed of where you are, just try to climb the ladder.

How to climb the ladder: metrics

To climb the ladder of proof, move the metrics and the numbers. A startup is a learning machine, and metrics show the team and the investors that you are learning and climbing the ladder. Everyone is watching your place on the ladder, probably including your girlfriend or boyfriend: "are you really going to keep doing this?" "Well yes, we just got revenue". "Oh, fine". Ideas in language, then you build something, then you have a product, then you measure it. Measurement is critical to proving upward movement. Without data there is no learning. You have to be obsessively focused on metrics, more than it seems. You have to decide what to measure. You can measure everything, but if you look at everything, it will get confusing. You have to think about what to do with the product and the system so that the metrics are timely and accurate. Because in the end investors will ask how you know this number is correct, and you will have to explain how you got it and that it is accurate. You have to think about how the team sees the metrics. The team can be investors, teammates, advisors. How does everyone see the metrics? How do we as a team make decisions based on them? And how do you show them to investors? And measurement and reporting systems take more engineering time than it seems. We see this flaw in many founders: they do not allocate enough time to how to measure accurate, timely and displayable metrics. They think they are just building a product, but that is only part of it.

As for display, you have to think about how to show the metrics. Is it a spreadsheet or a PowerPoint, that is the difference. Will it fit on a smartphone, on a picture you send that will be visible on a smartphone, or not? That is also about display.

One of the things the author always does with his companies: from day one he starts sending everyone on the team an email with a summary of daily statistics. At first it will be bad, there will be only two or three metrics, because the systems are not built yet. But this becomes embedded in the team culture: daily statistics are up and running, you add new metrics, the metrics become more accurate, people understand what to look at and what matters in their roles. Then this daily data can be tracked longitudinally, and it rolls up into what you show investors.

Another technique that is always used: the metrics are put up on the wall in real time. When you walk into your room and see a screen on the wall with the live statistics of the product, even if it is your sales department, because it is not necessarily consumer internet, it can be a sales funnel, any metrics you generate, it inspires and shows what to focus on. You can show any visitor what we are focused on, and people understand what they need to move to climb the ladder of proof. And always once a week they gather to project this on the wall and point and say how it is right now. This interface, this display is the display. You have to think about what display you use for the metrics. Many do not spend enough time on this.

Which metrics to measure? Only the metrics you measure move. Many say they want more customers, but do they measure that daily? No. You have to measure and show everyone, otherwise they literally will not move, no matter how you manage the startup. And you have to show it correctly. Focus on the metrics you actually want to move, and think about the ones investors will want to see growing, to climb the ladder.

Also, focus on metrics that change your behavior when they change. Many have vanity metrics that feel nice and let them sleep at night, but they do not change how you run the business or the product's behavior. Measure painful metrics, the ones that will force you to change behavior when they change. One of the most common metrics is the sales funnel and the conversion at each of its stages. This can be done in a spreadsheet or in a chart, but almost everyone should do it fairly soon, to build the habit. Another favorite metric is retention curves.

Once the author was asked what he does, without knowing he was the CEO, and he answered: "I am a reader of triangular charts". That is his main job: to read them every day, every week, to see how we are doing with retention, because retention is the source of long-term value. You have to build a system or pay for SaaS that builds such charts, so that you and the whole team can observe them, this is often a very important part of such businesses. Many come and ask which metric is important, and in eighty percent of cases the answer is retention.

The "North Star" metric

There is a concept of one metric: if you want everyone to be aligned, find one metric that matters. This is the North Star metric. It is needed to focus the team's attention and its resources, because it is always in front of people, the biggest number on your display: in the email, in the fundraising deck, on the screen in the room. This one number aligns everyone in the direction the company is moving. It should measure customer value, be actionable and directly correlate with revenue. This can be different for different businesses, but you have to figure out as early as possible what this metric is. For the first three months the North Star metric may differ from the next three months, and that is normal, you change them from time to time, like different innings in baseball. But you have to know what it is right now, and tell the team: this is our North Star for the moment. Do not say it is forever, otherwise they will ask why you changed the metric if you yourself do not know what you are doing. Say: right now it is this one, then there will be the next one. Examples: monthly revenue, the Net Promoter Score index, daily active users, retention figures, transactions on a marketplace. If you got a transaction, then the buyer and the seller were satisfied. There are a bunch of metrics for the seller and a bunch for the buyer, but your one North Star metric is whether we got liquidity on the marketplace. This is an example of a simplifying metric that was often used for marketplaces.

Milestones as rungs of the ladder

Another topic in conversations with investors is milestones: "I have reached such and such milestones". What are milestones? Milestones are simply movement up the ladder of proof, that is what they are. That is why people care so much about milestones. They reduce risk and prove that the business works. You have to determine on a table which future events will move you up the ladder of proof, which will reduce risk and raise the chances of success. Hiring a chief revenue officer, a lot of things need to be done. If you can write it down, everyone will be able to help you reach these milestones and climb the ladder. Write them down, make them actionable, measurable. And then ask yourself: how do I cut this number in half, to get to rapid growth, rapid revenue growth, that unique rung that eliminates all the others? How do I eliminate milestones to get to rapid revenue growth? And how do I cut the time in half, because you definitely can. The great work of an entrepreneur is to simplify, simplify, simplify. That is exactly why Elon was able to build rockets for one hundredth of what NASA spent: they simplified, eliminated milestones, removed steps, removed time.

Move the rocks first

To move the numbers up, you have to move the rocks first. You have to assemble the right team, all together. It does not matter whether it is in a dorm room or you meet every day for six hours in the cafeteria, get everyone together. Without being together, you will go much slower, and the numbers will not move. You have to get a story that makes sense to you. You have to get a name, which has already been discussed. You have to set your head: a commitment to speed, your own flexibility. These big rocks let the numbers grow. Many founders cannot move the numbers because they worry about how they look in someone's eyes, or they have not set their head. They decided in their own head: "I just raised two years of capital, I do not need to move the numbers for eighteen months". Their heads were set wrong. And you have to get a customer, determine who this ideal customer profile (ICP) is, and get a distribution channel. These are the main rocks for moving the numbers up. And when all the rocks are in place, iterate like the devil. Even if you are building something that will launch in a year or two, iterate like the devil in how you approach building. Become that learning machine and those experiments, and the results of the experiments are the metrics. What is stopping you from getting these metrics? You can move them much faster than it seems, and climb the rungs of the ladder of proof much faster.

Linking nodes: a startup is a network

In preparing to talk with investors, think about the concept of linking. You want to link the investor to your network. Your network is your startup. A startup is a network. You want to link investors to the startup, customers to the startup, journalists, recognition and influencers to the startup, talent to the startup. Once you start thinking this way, a lot will open up. Each node brings into your network its own existing network, the attention it already has, its status, traffic, capital, skills and credibility. Aggregating these resources is the essence of a startup.

How do you compensate each node for joining your network? To the investor you give shares. Shares are just a piece of paper on which you wrote "I founded a company", and you give it away, which is worth nothing right now, and he expects it to be worth more in the future. A good deal: he gives real money for your paper. With employees it is the same: you sell them dreams, maybe titles, maybe status. You should expect that more influential nodes, for example, more influential investors like Sequoia or Benchmark with greater credibility and status, will demand more compensation, in your case shares, to support your project, because they are lending you their credibility and status, which lets you aggregate more nodes. They know their worth, so they will expect more. Besides that, nodes that join early will demand more, because they know they are taking on more risk: you have not yet climbed the ladder of proof. So they will expect more in compensation, titles, shares, salaries. When you raise the first two million, maybe at a ten million valuation, you give away twenty percent. But later, when you have climbed the ladder, someone will agree to take eight percent for two million, someone four percent for two million, and eventually someone will take two percent for two million, and then maybe 0.02 percent later. This is what you can expect as you move up. The same is true for employees.

Not a pitch, but a conversation or a review

The author does not like the word "pitch". A pitch is when I throw the ball one way, and the ball flies right there, and it is not like that. It is better to think of it as a conversation or a review: we are looking at this together, or consulting: "I wanted to consult with you on these questions". The first five conversations with investors are needed to learn. Do not only try to convince, do not pitch, do not only try to sink the ball into the hole: "give me money". Start a conversation where you ask good questions.

One founder, whom the author has worked with for three years and who is raising a series B round of about forty million, said that a question the author told him about became his main question for everyone. If he has half an hour, he pitches for eight minutes, and then immediately asks this question. It is a "can opener" that lets him learn about the firm, about how the other person thinks, about how his pitch looks. And it lets the person say: "others will be very skeptical of you because of this, that and the other; not me, we have a great relationship, there will be no awkwardness, I am just telling you what others will say". This is a good disarming technique to move to an honest conversation. Every venture capitalist will like such a conversation and such a question: "what is the one thing you think I am underestimating or being naive about?" They love such an honest conversation. Another good question: "what other companies does this remind you of?" And: "are there patterns, positive or negative, that apply to this business?" For example, a pattern: doctors cannot be made to change processes and behavior at all, and this is a big brake on a business of this type. There are thirty six similar questions in total, they are written up in the article "The Winning Psychology of Fundraising for Founders" on nfx.com. These are the questions every investor will be open to if you do not burn all thirty minutes on a pitch. Pitch for eight minutes, and then ask questions. The old pitch model from the eighties and nineties was needed because there were only one hundred fifty venture capitalists, each was worth its weight in gold. Now there are thirty five thousand, meet three a day, they are plentiful, and you can learn a lot from each one, it is a huge resource for learning.

People say: "I do not want to ask these questions, or else they will not invest". The answer: they will not invest in you anyway. In the nineties you had to build a relationship with a venture capitalist as early as possible, because there were few of them and they needed to understand companies. Now Sequoia will hear about you on Thursday, meet on Friday, meet again on Sunday, ask you to present on Monday and issue a term sheet on Monday evening. NFX will do the same.

Module 8. The three-person unicorn

Episode 8 of 9 · team scale in the age of AI

The eighth episode does not claim that you should only build a three-person company. It uses this as a way to discuss the principles of speed. One of the big principles: big teams go slowly, and small ones go fast. This is a concept called allometric scaling, and it is the subject of this episode.

Allometric scaling

Allometric scaling is a biological term, and it means that if something becomes heavier, it needs to become even heavier to support that weight. And conversely: if something becomes lighter, if you have fewer people in the company, then the systems needed to support the lighter weight of the company also become lighter, and the company can become smaller and smaller. If you apply allometric scaling by building a company small, fast and dense, it can become even smaller, even denser and even faster. Fewer people means less drama, less politics, less payroll, less dilution, and you bring into the team only the A+ level people you want. You do not have to fill the staff with whoever you could find. If you plan for fewer people, then every operation of the business becomes simpler, faster and cheaper.

Billion-dollar companies with thirty people

In the age of AI this is even more true. Billion-dollar companies with thirty people or fewer are now appearing at a rapid pace. Two years ago the firm wrote the article "The three-person unicorn", predicting this, and here two years later it is reality. Cursor does about one hundred million in ARR with thirty people. ElevenLabs does about ninety million in ARR with fifty people. Mercor does about fifty million with roughly thirty people. Lovable reached ten million in ARR in two months with fifteen people. This is not the future, it is happening now, the numbers are in hand. This is happening far more often than ten years ago, when there were maybe Instagram, WhatsApp and Groupon, and only very narrow types of businesses could grow at that speed. Now many different types of companies will grow this way. These are teams that fit in a conference room.

Systems as spinning plates

The first thing to realize: you need people who can set up systems. You set up AI agents and systems for sales, marketing, customer service, HR, legal questions, all of this can already be done today. You set up these systems, and they are like spinning plates: you have to spin them up, and then periodically come back and check so they do not fall and break. The old rules about team size and how much can be done are gone. That is exactly why the need for people shrinks sharply.

Three roles: vision, numbers, words

You cannot just throw three people into a room. These people have to be able to manage AI systems, wear several hats and work at a very high level. Three types are observed. The first is the visionary founder, like Steve Jobs, who sees the opportunity before others and really drives the product, a product visionary. The second person is the numbers person, someone who thinks in numbers, metrics, loves spreadsheets. He runs finance, operations, loves details and does not allow decisions without data. Just like the Mentat from "Dune": everyone looks at him when there is a question, and he always has the answer. In the film he instantly names the exact cost of the flight: 1.46 million Solaris, plus 62 Solaris for the return trip, because the numbers person always keeps the figure in his head. The third person is the words person, Don Draper, if you like, someone who can turn the AI input into a human message, someone who gives that feeling, that taste, that brand, which will be absolutely essential in such a competitive world. Not everyone has to do everything, but everyone has to be able to manage AI systems so the plates keep spinning. Vision, numbers and words: that is what a three-person unicorn means.

Ninety percent agents, ten percent founders

What AI cannot do is have flashes of insight. That is the founder's advantage. Maybe ninety percent of your company will be AI agents, but ten percent will not, and that is where founders are needed, and that is why founders remain the most important part. On most days AI will outdo most people, but there are moments when insight is needed, a change of direction, a change of mind, a change of feeling, and only founders can provide them. The key is to act at those moments, and that is the founder's advantage. And the key to that is to get out of your own way. You have to unlearn what you learned in the past, so that this becomes the link between you, your few co-founders, your few employees and the AI. That is exactly how very small, very fast-growing and very large companies will be built in the future.

Two years ago it was predicted that there would be three-person unicorns. Even now companies with fifteen, ten, thirty people are visible, approaching that mark or past it. The next prediction is this: hyper-unicorns will appear, that is, unicorns that become unicorns just six months after launch. This goes beyond startups and penetrates everything, including content: the script for this episode was written by AI, the video was edited by AI, and the only real thing in this piece of content is the person in the room and the camera in front of him, or maybe it is an avatar.

Module 9. Hidden patterns of great ideas

Episode 9 of 9 · choosing an idea

The ninth episode opens with the strongest line of the course: the biggest waste in the entire startup ecosystem is great founders working on bad ideas. So we have to break down the hidden patterns of great ideas, in order to tune into the right direction, because if you are going fast, it does not matter if you are going the wrong way. Most things in the startup world do not work. The episode is meant to save years. The firm runs the FAST program at Stanford GSB.

The tragedy of talent on bad ideas

Last year the FAST program had seventy-five teams from the Stanford community that applied for a $100,000 check, a $250,000 check, an uncapped convertible note, all founder-friendly terms. They gave out money and got several great companies. However, out of the seventy-five companies, sixty-eight ideas were bad. The author simply knew they were bad, and the founders themselves did not know, because he is old and they are young and he had seen the patterns. They have naivety, talent, drive, and intelligence, everything they need, but most of them start in the wrong direction. After that the author told the NFX team and his fellows, of whom there are six on campus, that they had to intervene earlier, before the April applications, because people work on bad ideas for months. This is the biggest waste in the startup and venture ecosystem around the world: talented people working on bad ideas. It is a tragedy.

The spectrum from fertile soil to mediocrity

The course presents the landscape of ideas like this. On the right is fertile soil. If a team of two or three people comes with a good presentation and a good idea, and it is fertile soil, then if you work as hard as you can on it for two years, you will sell the company for ten million dollars or more. To the left of that is the "not sure" zone, and at the far left edge is mediocre, and in roughly sixty-five percent of cases the author can say that it will simply waste the energy of your life. In the middle he says "I am not sure." Airbnb was in the "not sure" zone: "will anyone sleep on an air mattress in your spare room?" I do not know. It turned out they would, but at the time no one knew, and the company turned out to be great. The author's job is to move you from left to right, and your job then is to execute.

This is where you should build big, world-changing companies, not small ones like McDonald's or a bakery. World-changing companies, as a rule, have network effects. The firm did a study: about seventy, more precisely seventy-two percent of all the value created in technology over the last thirty years came from companies with network effects at their core. You can build a corporate SaaS company and sell it for a billion, but it is hard to build and sell for ten billion. Those are still good businesses, but most of the returns in technology and most of the influence come from businesses with network effects. The firm discovered at least sixteen network effects and published them. So think about how to embed network effects into the company before it even launches. That is why the firm is called NFX, from network effects, it is that important to what makes a company great.

The three S's: simple, scalable, similar

Follow the three S's. First: find something simple. Many years ago the author heard the line that at Harvard Business School, those who got A's worked for those who got B's, and the C students made all the money. Why? Because the C students simplified, and business is simple. You as Stanford people have a disadvantage: you are too smart for business, you want to complicate everything. You complicated things in fourth grade, eighth grade, twelfth grade, college, now here, because you are smart. And school and university reward complexity: longer essays, deeper thinking. Business rewards simplicity. The author has repeatedly observed startups struggling with their four million until 400,000 is left, and then they say "to hell with it," reach the "to hell with it" moment and make the right decision. Why? Because they simplify. They are under pressure, backed against the wall, and finally start thinking clearly, stop complicating. Right now the author has three or four companies in his portfolio that he lets get to sixty days from the end of capital, he pushes them toward it, and they learn fast and simplify. You have to simplify as early as possible.

The second S is scalability. This is not about bakeries, but usually about software, biology, technology at scale. We are very lucky to have been born in this century: we are not making shoes in 1950, we are not farmers, we make things at scale. Think about what scales, especially on the distribution side. The third S is similarity. Most of the talk about entrepreneurship is about something brilliant and breakthrough, but if you look at the history of what created big companies, the largest of the known companies were very, very similar to companies that appeared just one, two, three, or four years before them. They were similar to something already working. Hence the motto: copy a lot, but not everything.

What this means in detail. You can imagine that your idea is completely new, and everyone will say: "oh man, that is so cool, so exciting, you absolutely have to do it." They will say that, but they will not buy your product, because they probably do not need it: something completely new is unlikely to be something people need. That is a fact, you will most likely fail. The other extreme is pure cloning, pixel for pixel, patent infringement. If you clone purely, they will come after you legally, and the community will say you are a scoundrel, everyone will say there is nothing original here, and you will feel your support falling away. So you cannot simply clone, at least in the US. If you want to clone, do it in another geography, that is normal, it is actually a good model. But usually you need to be in the "new energy" zone, where you make two or three obvious changes.

New energy: two or three changes

Examples. The company Prosper was building a lending platform, and two years later Lending Club essentially cloned it, changing just two things: they pooled the loans and talked to the FTC before launch. Those two differences made all the difference. Lending Club took off to roughly a seven billion dollar valuation, while Prosper was effectively shut down by the FTC. Later Prosper came back, copied what Lending Club did, and now they are about the same size. By changing two things, they got something completely different, and it worked. The same with Google and InfoSeek. The same with Airbnb and VRBO: a couple of changes to something very similar, and no one thought it was just VRBO and that these guys were scoundrels; everyone said Airbnb was its own thing. Facebook was simply a copy of Friendster, the same blue, the same five features. There were about a hundred social networks at the time before Facebook launched. It was very similar to something he knew worked. That was the core, the beginning. And then it captured the network effect, took the best team, and made its own. That is usually where this comes from, not from what is taught.

Market risk versus execution risk

Another important point is the distinction between market risk and execution risk. The author's first four companies carried both market and execution risk at the same time, and he calls this stupidity. Elon Musk tried it with x.com, which merged into PayPal, and after that he said he would never take market risk again, because who knew whether anyone would want to send money over the internet, that is exactly where he started. Huge market risk. So he no longer takes market risk: he knows people drive cars, he knows people pay money to send things into space, and he knows people put solar panels on their houses. No market risk, he takes only execution risk. Very smart. First-time founders, as the author once was, often take a lot of market and execution risk, but since they have never executed, it is unlikely they will do it well. Second-time founders, who know they execute well because they failed many times and learned, are smarter: they take only execution risk, because they know they are good at it, they can raise money, hire people, they know the right balances, they can move, maneuver, but they do not want to take market risk. It is too big, too unlikely to work.

Another view of risk came from Rick Frisbie, the founder of Battery Ventures, when the author worked there. The author told him they should invest internationally, that all the hot markets and big returns were there. Frisbie replied that he did not think so. The author insisted, and he and another partner made two investments and within a year lost a hundred percent of the money. Four months later Frisbie came up in the hallway and said he had figured out what happened: when you invest in an American company, there are, say, eight big risks: execution, market, legal, product, and so on. When you invest internationally, there is a whole additional set of risks. So systematically, a venture portfolio with international investments will show lower returns than a domestic one with less risk. The author spent eight million dollars to prove this, and passes it on. And it is not only about internationality: think about everything. What if your market is regulated? There are markets without regulation. What if you have to roll out city by city? There are markets where you do not have to roll out city by city. What are all the risks you are taking to make your business work? Write them down and compare them with other ideas. It is not that risk is bad, we are in the business of taking risk and doing what others will not, but you have to understand your risks and map them before the start. For example, in medical software the level of risk is so long, the barriers are so high, and the history of losses over the last fifteen years is so deep, that you have to be very sure you want to spend the energy of your life on it for the next ten years, because getting through it is very thorny, there are so many risks and blockers.

Red ocean versus blue

Usually people say: what if Google does it, and Sequoia has already invested in a company in this space, it is a red ocean, do not go there, it is dangerous, too much competition. But going into a blue ocean, where no one is, is also quite risky, because you take more market risk. It has been repeatedly observed that people who go into red oceans end up doing better than those who go into blue ones. The author has a company that tried two blue oceans and still has capital, so it is trying a third. He says: find a business you admire, and find the new energy around it, start something very similar, because you know someone will buy it, you know how to value it, you know the go-to-market, because you see what the competitor does. Enter the business, let blood into the system and let the system work on the blood, because right now you have no blood, you are too far out in the blue ocean. Enter the red ocean, get revenue, get customers, and then learn what is needed next. Then the whole organization can look at revenue, at sales cycles, they will have real numbers, and they will build the company. Maybe it will not be a billion dollar company, but maybe at least they will have a company. The firm ran an accelerator with seventy-five companies, and eight of them are already unicorns: the methodology works, it is proven in practice.

Timing: platform, cultural, and regulatory shifts

How do you know the timing is right? Usually we do not know, no one really knows, but there are signs. Usually it is rapid platform shifts in technology. This is different from a specific technology: platform shifts are when something suddenly became possible. And it does not mean slow technological shifts, but rapid ones, like the one that happened with generative technologies over the last eight months. Generative technologies had been around for years, the firm had invested in them for years, but suddenly the price fell by a hundred times, and it really began to feel like a person writing text, and like an artist painting a picture, which was not the case two years ago. There was an explosion, a Cambrian explosion of generative technologies. This is a good example of a platform shift, as are mobile phones, blockchain, or certain types of fundamental technologies. When you see that, it is the key that something became possible for the first time, that is exactly what you look for in terms of timing.

Many say that generation Z buys differently, generation Y feels differently. But these are things that unfold over a decade, which means the whole competition sees them the same way and can take advantage of them. You will not find truly open opportunities by talking about cultural shifts. People confuse this, thinking that this is what changed, but no.

Sometimes there are regulatory shifts. A few years ago the regulation of how the state pays for dialysis changed, because a new technology for home dialysis appeared. The regulation was changed after ten years of lobbying, and when it changed, over the next three years an opportunity opened up for new dialysis companies. This is also a good moment for a "why now", but if you had tried to do it before that, it would have been bad. Many come and say there is a law that will change next year. No, it will not change, it will be postponed to next year and the next, and the next administration, and it will come in four years. Do not try to anticipate regulatory changes. The opportunity in dialysis became available another four years after the law changed, because no one really grasped it. With regulation, do not try to be ahead of it, be after it. And with technological shifts, you want to be right there.

The second point of timing: the market is growing fast. Rapid new behaviors, for example home stays instead of hotels, became a thing very quickly, and many big companies were and continue to be built around this change in consumer behavior, including Airbnb. Many ask what your TAM is. The TAM is small today but will become big. You just need to see a fast-moving market, even if it is small now, a big TAM is not required. And you should not always say: the transactions that happen on this B2B marketplace, we will move them into digital. We tried, it usually does not work. What is needed is to find new transactions that do not happen, in a subsegment that may end up taking most of the market, and attack there. Create new transactions, do not try to replace old ones just because the TAM is big. Do not be fooled by big TAMs alone.

And also: it is not a platform shift if it is some specific thing from your lab. Many come and say: I made this in the lab, I want to turn it into a business. This rarely works. The phrase that has been heard in the venture business for thirty years is "technology in search of a market." This is different from a platform shift, it is just a specific thing you may have discovered, and such things rarely turn into anything, unless they themselves are platforms, and then you have several attempts on that platform. That is why in biotech the firm mostly invests in platforms, because it may not work for the first six indications, but work for the next two, and then you have a real company.

Boundary layer and area under the curve

Another big pattern that the firm spent a long time figuring out. A few years ago there was a company Wonderhill, and it was building video games. The first few games did not have enough users in terms of the number of potential customers, and people used them too infrequently, and the average selling price, the amount earned per day, was insufficient for a business. This was inside a certain boundary layer that prevented it from becoming a real business. Then they looked at a company that was successful at the time, FarmVille, and it was earning four cents a day from roughly a hundred million people. They had a real business, they advertised, they had staff, they were growing. Everyone talked about it. "Can we build one like that?" They tried but fell short, got to about thirty million, and that was not enough. Then they built the game Dragons of Atlantis, similar to another game, Kingdoms of Camelot. It had fifty thousand users, but they earned almost four dollars a day, and that became a real business: there was staff, things were going well, it was great, because it was outside the boundary layer. So when thinking about a business, think about how many people you can realistically get.

Someone came with generative technology for 3D modeling for game developers. The author asked: okay, and how many game developers exist? Surely not fewer than ten thousand, maybe fifty thousand. There are many people, but is that enough to build a truly big company, or will it become an acquisition for fifty to a hundred million dollars? And indeed, all 3D modeling companies over the last ten years were bought in the range of fifty to a hundred million dollars. That is great: if you want to start this, you can be in the gaming industry and make ten or six million, that is amazing, but it is not what the author strives for.

And the game Mafia Wars was actually the game that drove Zynga: it had about eighty million users and about two dollars and seventy cents per user per day, and it drove them for two and a half years. If you want to build a video game, think "Mafia", "Pirates", "Mafia", "Atlantis": that is what works. In fact, if you want to know what works in games, go to the Lego section of a toy store: they have "Pirates", "Atlantis", "City Builder", everything that works in video games. This is another pattern. The reason Mafia Wars works is that the area under the curve is so big. The area under your curve must be big, and you must be outside the boundary layer. This is ASP versus market size.

Pivot, name change, and the skateboard

When to pivot? When to hold on? It is very hard to know, usually it is also unknown. But it is clear that you can pivot and still build a decacorn. Many have done so, because they had passionate convictions that are easy to hold: they truly believed in what they were working on, but when they got data, they easily let go and moved on, and taught the team to feel the same way. Sometimes the hardest thing is to teach the team to let go easily, because they are trying to do a good job for you. There are many companies that pivoted and became big and important, but you cannot hold on too hard, and you must be open to pivots. Of the author's four companies, three pivoted and all were successful, and in two of the four they changed the name, because after finding product-market fit they realized the name did not fit. Do not be afraid to go down to the very foundation, to the name of the company, and start over, it works.

As for mindset, there is a trap. Because you are successful, you have been told you are the best: the best in school, in high school, in college, here. And it seems to you that everything worthy of your effort must be grandiose. You are in love with the sports car. You get up in the morning and say you are building a sports car, and everyone says how cool you are. But you cannot build a sports car right away, it usually does not work that way. You cannot fall in love with it. You have to be ready to build a skateboard, ride it, and say you are building a skateboard, that you are thrilled with the skateboard, that it is the best skateboard. And they will tell you it is not that cool, they want to save the planet, to work on climate tech. And you have to be ready to put your heart and soul into the skateboard, because once you have built the skateboard, you can move up toward the sports car. Sharing photos in college, who thought that was a big idea? What you ate for lunch in a hundred and forty characters, was that a big idea? Renting an air mattress in someone else's room, was that a big idea? Almost every company you admire, except maybe Snowflake, started with something very small. Do not indulge your grand plans about how cool you will look because you are building a sports car. Be a little more modest.

Commando, playing to win, and an ambitious vision

Another indicator of where great ideas come from is whether the idea attracts commandos. There is a commando mindset that will take the hill and do whatever is needed, and there is the soldier: setting up white mobile homes, laying dirt roads, one light bulb at the pontoon bridge, and things under control. And there is the police officer, whose only job is to prevent anything bad from happening. If you are a police officer or a soldier, you are unlikely to succeed as a startup founder, very unlikely. Most venture capitalists are "sniffing out" this, because the track record of soldiers and police officers shows that they are actually trying to do such startups. We always sniff out these personality types, because they are deadly. If you have a soldier or a police officer on your team, you have to remove them. It is very hard to change this mindset, people's fear level is set at a certain place and it is hard to shift. One of the reasons Facebook was so good up to about 2014 is that they kept the commando culture. Everyone there stayed commando up to about that time, which is why they won again and again. Even at scale, when they had 20,000 people, the culture remained commando. So do not say you cannot find eight commandos for the campaign, you can. One of the reasons many ecosystems, like Atlanta, Orlando, or Cleveland, find it hard to produce big technology companies is that few commandos remain there: they left, they did not feel comfortable there. You need a commando mindset to do something great, fast, on the edge, something people smirk at, because soldiers and police officers do not like being smirked at, especially the smart ones.

Another mindset is playing to win versus playing not to lose. The author was one who played not to lose. That is why all four of his companies turned out "fine", but he never built a billion dollar company. His partner Pete Flint built one, another partner "Giggey" built one. The realization came only with the third company: "oh, this is all going well, but we keep selling for a hundred and fifty million, what is this?" It is because of the mindset. It harms you, it stops you from doing something great. You have to move into the space of playing to win, where you are ready to take risk, to strike, and to really attack something, when you already have product-market fit, when you are ready to go all in.

An ambitious vision seems like a platitude, but it is surprising how many people do not have it. If you have this mentality, it is contagious: it attracts the best people, the best capital, the best ideas, and people come and give you their best ideas, because you can do something with them. You have to surround yourself with people who support an ambitious vision and push you. Some guys raised six million in Toronto, and four days a week they were asked to come to breakfasts and give lectures on entrepreneurship at the Holiday Inn. They were the kings of Toronto. But if you are standing at a football game here in Palo Alto, and you raised fifty, and the guy next to you raised two hundred, you just say "oh, I need to keep going." When you are surrounded by a big ambitious vision, your ambition grows.

Recruiting the best, the power of people, and company mafias

Another pattern in the minds of those who build the largest companies is the ability and desire to hire absolutely the best talent. Facebook: there were a hundred companies with the same five features, but it had the best team. The author had a team with a social network of thirty million people when Facebook was just launching, but Facebook had a better team: better design, faster-loading pages, real names, they iterated faster, and the author was also very fast and was among the best product designers in the world, but these guys were on another level. They started with a network effect, hired the best, and boom. Microsoft started with a network effect with its operating system, they got lucky, they did not know what it was, they stumbled onto it. But they recruited the very best people, their hiring process simply sucked talent from all over the world, brought them to Bellevue, and they crushed everyone. The same with Google: they copied InfoSeek, they had an algorithm about ten percent better than AltaVista's, but the story that they had better search is not true. AltaVista was better, about the same. They had better people. Eighteen months after their first funding they were better, but at the moment of launch they were not better. The idea of finding and hiring the best people, who are better than you, again and again turns out to be a winning combination, especially when most of your businesses are new energy in a red ocean.

To feel the power of the power law, watch the YouTube video "These Regular Guys Challenging NBA Player and Instantly Regretted It." They find a forty-two-year-old man who played in the NBA from age twenty-seven to thirty-three and was three hundred and sixty-third out of three hundred and sixty-five in the stats, that is, the worst in the NBA for three years. Now he is twelve years older, thirty pounds heavier, and they take the best from the G-League and the best basketball players from everywhere else and play one on one. He just destroys them with scores of thirteen to zero and thirteen to three, eleven to three. And you realize: wow, he is the worst of these, he is here on the power curve, and he is so much better than everyone else. When you feel this power law, you will become more committed to hiring the very best and to working on yourself to become one of those people. You do not want to stay in the G-League, you want to be in the NBA. For that you have to work hard on yourself: on authenticity, on communication, on the willingness to admit you are wrong, on charisma, to pull people into the team, on energy, to draw in all the nodes and connect them to your network, which is your company.

There are centers of excellence and "company mafias": many great companies and founders came out of Facebook, PayPal, Stripe, Square, Salesforce, PeopleSoft, and unit 8200. There is a pattern of geography. This is a controversial thing that no one likes to talk about. There are articles "Your Life is Driven by Network Effects" and "Where To Build Your Startup". In short, the Bay Area is seven times better than any other ecosystem, at least seven times. There are many reasons. One is that you can hire a really good sixty-eighth employee, and in Atlanta that is very hard to do. Another is that you get really good advice from people covered in scars: they tried what you are doing, made a bunch of mistakes, and now they will warn you, save you a week, a month. Another wing is the exchange of KPIs. The author recalls how in two of his companies, when they were wandering in search of product-market fit, a conversation with someone gave a key metric that changed the business: suddenly they became profitable, or the business doubled or tripled, because the way of sending emails allowed seven percent of clicks instead of two, and suddenly there was a business. These are small details. The same in creative industries like film in Los Angeles: how to turn the Fresnel lens so the light falls on the face that way; do not do it that way, because it looks like television, and you want to look like film. These small details are needed to reach the level of excellence. When you are in such an ecosystem, such information flows to you. They try to make this work online, on Zoom, but it is not the same. That is why everyone says "back to the office", especially in creative industries and especially when competition is high. This will continue for the next four years, until everyone says they are back in the office. If you want to do something great, you have to be with each other.

Idea-finding methods and a critique of customer discovery

Idea-finding methods are where the author becomes controversial. How many great companies that we admire came from a process of interviewing customers and asking what they want? These methods are sold for consulting fees to the US government and big corporations. Do not go down that road if you can avoid it, look for other ways to find an idea. Why does this not work? Psychologically, because customers do not really see the future. It is you who must see the future. At one pole is Steve Jobs, who says: to hell with the customer, they do not know what they want, I will tell them what they want. That is an extreme, and the author does not preach it, because very few are that talented. But he preaches roughly this, rather than going and interviewing customers, asking what they want. Recall the story: if you had come to the docks in the 1800s, where people carried sacks of grain up the gangplank into the boat and dumped them into the hold, and asked "what do you want?", what would they have answered? "I need better boots." No one would have said: I need a piece of metal with a block, to lift this onto a pallet, carry it, and dump it. No one would have said that, they would have wanted better boots. So you have to say: I will take a block, take a bearing, and do it differently. "Well, that will leave me without a job." "Yes, it will leave you without a job, but now you will not have to carry this on your back, you can go into paperwork or become a banker." So breakthroughs also come from those who do not use the product. And again: do not chase the main transactions on a B2B marketplace, but go after the side ones, after the non-users, create new transactions.

This leads to an important point: start with the transaction. If you think you will start a business around a certain transaction with a certain person, that is the right place to start. The author says that if you want to really change the world, over the last twenty years the greatest difference he can imagine was made by guys from Scotland. They bought up used European GSM mobile networks that were being replaced, brought them to Africa, and set them up as purely commercial, profitable businesses across Africa, and suddenly everyone could do their own thing, make their own decisions, and live their own life. This transformed economies very positively, was purely commercial, and laid the foundation for everything else. It does not look like missionary work, but it had a huge positive social impact. And now everything has evolved into having to be a missionary founder. No, first you have to be a fast founder, a "wild" founder, a competitive founder. And only then can you have your mission and fulfill it, when you already have a business that is going.

Cross-cutting Section A. How the four levers interlock

The nine modules go through the material episode by episode, but the course rests on four levers: speed, language, idea, and psychology. Understanding how they interlock is the real conclusion of the course. None of the levers works alone, and speed is presented as the main one, because it amplifies the others.

Let us start with the idea. If the direction is wrong, speed only brings you to the abyss faster. So choosing an idea, examined in module nine, is the filter without which everything else loses meaning. The idea sets where to apply force. But a good idea by itself is worth nothing without execution, and execution is measured by speed.

Speed is the master lever, because it accelerates learning. The more completed cycles you do, the faster you learn what you need to learn, and the faster you correct the idea. Speed improves the product, attracts resources, raises the valuation, and speeds up the climb up the ladder of proof. Psychology determines whether your own brain will let you go fast: fear, self-sabotage, the need to keep options open, the desire to prove you are right, all of these reduce speed. So psychology is what opens or closes access to speed.

Language sits at the center, because it shapes both the idea and the speed, and the psychology of the team. The right words change the TAM and the thinking about the product, which means they change the idea too. The language of buttons and positioning directly affects conversion, and conversion is revenue and the climb up the ladder of proof. Language inside the company shapes culture, and culture determines whether the team moves fast. Changing the wording can remove the fear of launching and thereby speed up the cycles.

Hence the picture: the idea gives direction, language gives form and meaning, psychology gives permission to move, and speed turns all of it into actual progress. If you remove any of the four, the system breaks. Speed without an idea is fast into a wall. An idea without language is a product no one understands. Language without speed is a beautiful description of something that is not moving. And all of it together without healthy psychology runs into the founder's fear and self-sabotage.

Cross-cutting Section B. Glossary of course terms

Below are the course's own terms, so they can be found and reviewed separately from the context.

Cross-cutting Section C. Summary table of all numbers and cases

Below, all the numbers, cases, companies, and names mentioned in the course are gathered into one table, with the source module indicated. The table serves as a reference for review.

Cross-cutting Section D. Contradictions, pitfalls, and limits of applicability

An honest training file is obliged to warn about the weak points of the source. Below are the caveats the reader should keep in mind throughout the course.

Contradictions within the corpus

Pitfalls and limits

• Self-reporting. All the material comes from the author's and his firm's spoken course. This is a self-report, not independent research. Many numbers are rounded in speech, many stories are told from memory. They should be used as illustrations, not as verified facts.
• Survivorship. Success stories are selected by the fact of success. The cases of those who did the same thing and failed do not make it into the course. Conclusions from survivor stories should be applied cautiously.
• American-centricity. A significant part of the examples, and of the advice on jurisdiction, hiring, and firing, is tied to the US and Silicon Valley. A direct transfer to another environment requires adjustments.
• Absolutism. The course likes strong statements: "remote work kills speed", "hire only commandos", "do it vanilla", "never do partnerships". They are useful as calibration, but as absolute rules they can be harmful in specific conditions.
• The venture capitalist's interest. The author speaks from the position of a firm that earns from the growth speed of portfolio companies. Speed benefits the investor; for a founder, in some cases it may be wiser to slow down.
• The gap between idea and execution. The course acknowledges that a good idea without execution is empty, but the main attention is given to speed and language, not to operational excellence over the long horizon.
• The source's prior. Where the course criticizes "customer discovery", it itself relies on selective stories and debatable generalizations. A critical attitude should be applied to the course itself, and not only to its objects.

Cross-cutting Section E. Final self-check: twenty questions

If the reader can answer all twenty questions without looking at the file, they have mastered the course.

---

END OF TRAINING FILE · NFX Speed Masterclass, nine episodes, retold as a course