Thoughts on Incubators

We have seen a huge rise in startup incubators recently. In fact, some people suspect that there are now more incubators than startups. The incubator hype followed the huge success of Y Combinator, pioneer and godfather of all incubators. I think that most incubators are destined to fail, and I believe that there are only two extremes of the incubator model that make sense.

Here are the two models that are able to win:

The accelerator

That's the Y Combinator model, the one that has been replicated successfully by other accelerator programs such as TechStars. Accelerators are beautiful things. I believe that they create huge value in early-stage inexperienced startups teams. That is because joining a startup accelerator is a magnificent replacement for business school.

In fact, Y Combinator calls a part of its program Startup School. It's better than business school. It's the hands-on curriculum that a startup founder should have. It's the condensed wisdom that helps avoid the most common mistakes that kill startups. What accelerators accelerate is the learning curve of the participating first-time entrepreneurs.

The accelerator's deal is the following:

You join a program for a well-defined limited period of time (usually between three and six months) and within that limited period of time, you get pushed to the limit. You get access to everything that helps you get your startup off the ground in a way an experienced entrepreneur would do it. They provide you with resources, knowledge, and their network. After the program, you can see the horizon, and the accelerator says goodbye. In turn, the accelerator keeps a small stake in your company. It's a fair deal.

The company builder

That is almost the opposite model. In my opinion, there is a reason why company builders aren't called "startup builders". It's because they don't build startups. They build model companies with proven business models. Company builders make a lot of sense when they focus on execution and execution alone. Successful company builders concentrate on a clearly defined market segment or industry, such as e-commerce.

One of Berlin's most (infamously) known company builders is Rocket Internet. Some people call it a "clone factory", a nickname referring simultaneously to their company building strategy (taking business models that have been successful elsewhere and replicating them in untapped markets) and their own company culture (a factory-like adrenaline-driven culture of execution).

The company builder's deal is the following:

You become part of a machine designed to rush pre-defined business models from zero to a hundred as fast as possible. The marriage is carved in stone and the company builder will stay until the company is profitable and ready for sale. The aim is an exit at a high valuation which will help the company builder refinance its operations. The company builder steers and controls the ship, the founder is just another captain. In almost all cases, the company builder owns a clear majority stake in the company and manages the company's fundraising. It's a fair deal for those who like it.

Mixing paradigms

Both accelerators and company builders can be successful. As soon as you enter the middle ground, however, I believe that things get shaky. Accelerators accelerate startups, company builders build companies. There is no way to "build lots of startups" out of a central incubator trying to take too much ownership and staying for too long. Conversely, you can't "accelerate lots of companies" out of a central incubator trying to establish highly effective processes and operations in just six months for a bit of equity.

Mixing the two paradigms doesn't work. It's either an autonomous founding team with an awesome idea and a clear vision, trying to change the world, dedicated for the long term, ready to starve and turn their dreams into a reality, no matter how long it takes. Or it's an execution-driven management team, chasing the money, leveraging shared resources and synergies to execute even faster and on an even larger scale to rake in even more money.

Two games. Two kinds of players. Don't mix them.


Great Things Aren't Obvious


What an interesting word. We rarely say "unordinary" or "not ordinary". We say "extraordinary". The choice is not between doing ordinary things and doing unordinary things. The things we're doing are ordinary by default, and we cannot stop doing ordinary things. We do them all the time.

The choice is not to be found in the things that surround us. Following our perception is an ordinary choice. The choice is in the conscious step out of the ordinary, into the unknown, in what is beyond, in the extra-ordinary. That choice can only be found in yourself, and yourself alone.

Doing great things

Great things cannot be done by copying what other people do. They cannot be attained by merely competing with others. Competition is another obvious choice. Living life with the goal of doing something a little bit better than everyone else is the Athlete's Choice. Athletes impress, but they don't create. The Maker's Choice is to do the things nobody else is doing.

The greatest companies are the companies nobody else is building. The greatest inventions are the inventions nobody else is thinking about. The greatest technological improvements are the ones that shift paradigms and reinvent a piece of the world. They are the unthinkable innovations.

Great things aren't obvious.


From Consumption to Investment

A while ago, I talked about extractors and enablers.

Today, I would like to explain the same dichotomy in different terms. In order to do that, I'm going to borrow some vocabulary from Umair Haque. According to Umair, economies historically develop through three stages: 

  1. Functional economy
    Survival & subsistence
    Making ends meet.
  2. Aspirational economy
    Consumption & opulence
    More. Bigger. Faster. Cheaper. Now.
  3. Meaningful economy
    Investment & eudaimonia
    Making people's lives better in real human terms.

During the industrial revolution, western civilization quantum-leaped from a functional economy to an aspirational economy. The problem is that it got stuck at that stage. Growing the aspirational economy out of proportion, we are now left with a fragile bubble economy. Global consumption skyrocketed since the 1970s, but global savings declined. Companies produce "razors with five blades instead of four" and "designer diapers". Meaningful growth has stagnated. Our economy needs a reboot. Things need to change.

The next step up the ladder is moving towards a meaningful economy. That economy isn't driven by consumption, but instead by investment. It measures value and wealth in different terms, using different measures. It acknowledges that today's economic institutions are broken, and puts new institutions in their place. Let me put the two worlds side by side of one another and appreciate the dramatic difference:

Consumption — the Extractor's Game 

  • Values:
    – Quantity ("more")
    – Size ("bigger")
    – Speed ("faster")
    – Price ("cheaper")

  • Paradigm:
    Opulence ("life full of stuff")

  • Driver:

  • Culture:
    Pay-backward ("paying back the debt")

  • Measure:
    Income as gross domestic product
    (flow of financial capital)

  • Nature of game:
    Zero-sum or negative-sum game

  • Stage of economy:
    Aspirational economy 

  • Characteristics:
    – Economic bubbles
    – Super-sized meals
    – Fast food and fast fashion 
    – Meaningless interactions 

Investment — the Enabler's Game

  • Values:
    – Quality ("better")
    – Relevance ("closer")
    – Sustainability ("longer")
    – Value ("richer")

  • Paradigm:
    Eudaimonia ("meaningful life")

  • Driver:

  • Culture:
    Pay-forward ("investing in the future") 

  • Measure:
    Outcomes in social indicators
    (stock of human and social capital)

  • Nature of game:
    Positive-sum game 

  • Stage of economy: 
    Meaningful economy 

  • Characteristics:
    – Technological innovation 
    – Enduring accomplishments
    – Healthy people and communities 
    – Relationships that matter 

Think about your life. Are you extracting value, or are you enabling value creation? Are you consuming or are you investing? Are you craving opulence, or are you making meaning? The day you look back on your life, will you have lived in your life, or will you have shaped life?


The Lean Idea

"I know that I know nothing."

— Socrates 

Lean is now officially overhyped.

Time to rediscover the essence of Lean. It's not about little capital expenditure and it's not about small teams. Most importantly, it's not about building crappy products. In reality, it's all about rapid feedback.

Incomplete information

Lean operates under the assumption that we have very limited knowledge about the rules of the games we are playing. It's a strategy deliberately designed to help us make the best possible decisions in a world of incomplete information. It emphasizes rapid feedback and an iterative process in order to counterbalance the penalty of missing information.

It's not a new idea either. Lean is the most humble answer to any big question. It is guided by the Socratic principle of the limits of our own knowledge, and the rejection of the Sophist notion of wisdom as an available resource. As a consequence of acknowledging our inability to come up with a perfect plan in advance, it replaces exhaustive planning with rapid learning and iterative improvement. "Too big to fail" becomes "Fail fast".

Lean is not bound to a specific discipline. Let me give you two very different examples illustrating the idea of Lean from supposedly opposing domains:

Lean in software engineering

In software engineering, we use design patterns to help us write better structured code faster. We use agile development in order to get to working software faster. However, I feel that the idea of Lean is most vividly expressed in the practice known as refactoring.

Refactorings are changes in the code made in order to improve its quality and maintainability without changing its functionality and without adding features. Refactorings should occur frequently and continuously. They can be thought of as the end of a feedback loop designed to test whether a certain piece of code can be implemented and whether its implementation delivers satisfactory results. Interestingly, the very practice of refactoring presupposes that the ideal code cannot be planned and designed in advance.

If the implementation doesn't deliver what was expected from it, throw it away—it's waste after all. If it does, however, refactor it in order to improve its quality and long-term maintainability. How to do things the right way can only be known in retrospect. It cannot be known in advance.

Lean in social engineering 

Social engineering is a term popularized by Karl Popper in his work "The Open Society and Its Enemies". (I'm not talking about the other kind of social engineering in the context of security here.) It essentially denotes the practice of designing and building a better society.

Popper distinguishes two fundamentelly different kinds of social engineering: Utopian social engineering and piecemeal social engineering. The former emphasizes long-term planning on a clean canvas whereas the latter emphasizes continuous learning and iterative improvement.

While both Utopian and piecemeal social engineers may initially have the same brave motivations, the former tend to bring about oppressive ideologies and horrible atrocities. By insisting in knowing the truth in advance, Utopian engineers gradually turn into enemies of the societies they set out to improve. Piecemeal engineers, on the other hand, are always concerned with removing the greatest and most urgent evil at any given moment, improving society as it is—iteratively, one step at a time.

As much as I love the theory of the Lean Startup, let's recognize the fact that its basic premises aren't that new. Iterative improvement and rapid-feedback learning are at the core of methodologies in various disciplines across the board. That doesn't make the findings less valuable.

It makes them proven, and durable.


The Startup Scale

What many people don't understand about startups is that one startup is dramatically different from another. When you think about joining a startup, the decision of joining one versus another is destined to have overwhelming consequences for your life. What do I mean by that?

Power-law functions and startups

Don't think of startups as evenly distributed groups of people. There is no "average" startup. There is no "typical" startup. The most "average" startup to be found is a failed startup. If startups work, they can work incredibly well. In fact, a successful startup will usually be many orders of magnitudes more successful than any other startup.

Startups follow a power-law curve and should not be measured like normally distributed quantities such as body height or life expectancy. Startups need to be measured on what I like to call the startup scale. The startup scale, like the Richter scale for measuring earthquakes, is a logarithmic scale. Very few startups make it to the top of the scale, and almost all startups we know — across the board — will end up at the bottom of the scale.

The reason is the following: It is as hard to grow a startup from $10 million to $100 million as it is to grow it from $1 million to $10 million. And that is as hard as it is to grow it from $100 million to $1 billion. Or from $1 billion to $10 billion. Or from $10 billion to $100 billion. For every order of magnitude of growth, the challenge is as difficult, and the ratio of startups who fail the challenge at that order of magnitude is the same.

Note: I'm using monetary valuation here as the measure of startup success. Obviously, that is neither the only nor the best measure. Any other measure will do just as fine. Should we choose "improving a certain number of people's lives" as a measure, for instance, the same logic would apply.

Founding versus joining a startup

The startup scale implies that choosing the right startup becomes critical. Only very few are truly world-changing, and most will remain mediocre. You cannot A/B test your way to success by randomly joining startups. You need to actively look for the most promising opportunity, the one where all stars magically align. Is the startup truly awesome, outstandingly ambitious? Is it going to convince every talented person in its vicinity to join?

Being able to move up on the startup scale a number of times is what matters most. Your initial stake, in turn, is not the most important thing. If you can join the Next Big Thing™, that is undoutedly better than starting your own insignificant business. A one-percent share in a $500-million company is worth twice as much as a fifty-percent share in a $5-million company.

The range of outcomes across startups is much, much larger than the range of outcomes internally within a startup. The janitor of the next Facebook will likely be better off than the CEO of your average inexperienced startup team. Understanding the nature of the startup scale is critical in order to avoid wasting time on startups that will never go anywhere.

Look for the most inspiring idea. The highest potential. The most compelling vision. The smartest team. The most beautiful product. The greatest culture. The most meaningful brand. Actively seek the organization you believe has the greatest chance to move up the startup scale many times in a row, and all you will need by then is to be a part.