.png)
Most founders learn about the venture capital power law the same way you learn a stove is hot: by touching it. They take the money, sign the term sheet, and only later discover that the fund's entire business model quietly assumed their company was probably going to fail. Not because the founders were bad. Because the math required it.
If you are going to seek venture capital, or accept it when it comes knocking, you owe it to yourself to understand the power law first. It is the single most important idea governing how your investors think, how they will treat you, and what "success" even means to the person on the other side of the table. Once you see it clearly, you can decide whether you actually want to be in that pool at all. In the agentic AI era, more and more founders are deciding the answer is no.
In a normal distribution, outcomes cluster around an average. Think human height. Most people are near the middle, and the extremes are rare and not very extreme. Venture returns do not work like that at all.
Venture returns follow a power law, which is a fancy way of saying a tiny number of investments produce almost all of the gains, and everything else clusters near zero. Out of a portfolio of roughly 30 companies, one or two will typically generate 80 to 90 percent of the entire fund's returns. A handful more might return capital or post a modest gain. The remaining twenty-plus return little or nothing. Horsley Bridge, analyzing thousands of venture investments, found that about 6 percent of deals drove 60 percent of the returns. AngelList's data tells a similar story: the best outcomes are not a little better than the rest, they are wildly, disproportionately better.
There is effectively no meaningful "average" outcome in venture. The mean is dragged so hard by the outliers at the top that it stops describing anything real. The fund does not need most of its bets to work. It needs a few to go to the moon.
Here is the part founders miss. The power law is not a flaw in venture capital. It is the design. A good VC builds a portfolio precisely so that one monster outcome can pay for all the losers and still deliver a fund-returning result. That means a VC is not optimizing for your company to succeed. They are optimizing for the possibility that your company becomes the one that returns the whole fund.
Those are not the same goal, and the gap between them is where founders get hurt.
If your startup could realistically sell for 30 million dollars in three years, that is a life-changing, generational outcome for you and your family. To a fund with a 400 million dollar pool that needs a few 50x outcomes to survive, a clean 30 million dollar exit is a rounding error. Worse, it is a distraction. So they will push you to swing for a billion, to raise more, to spend faster, to refuse the safe and wonderful exit in pursuit of the improbable enormous one. From their seat, encouraging you to risk the 30 million in hopes of the 3 billion is perfectly rational. They have 29 other bets. You have one life.
The same misalignment quietly bites angels too. A lot of angels do not actually run a 30-company portfolio. They wrote five checks to founders they liked. The power law is unforgiving at small sample sizes. Without enough shots on goal to catch an outlier, an angel can do everything right, back genuinely good companies, and still lose money in aggregate because they never caught the one that pays for the rest. The structure that works for a disciplined fund can quietly punish the individual who is only loosely playing the same game.
There is a second layer of misalignment that almost nobody explains to founders, and it has nothing to do with your eventual exit. It is about how a VC fund raises its next fund.
A fund's track record, the number it shows prospective limited partners to raise Fund III and Fund IV, is built largely on paper. Years before any company is actually sold, the fund reports its performance as IRR, and that IRR is driven by the marked value of its portfolio. The way you mark up a private company is simple: it just raised a new round at a higher price. A step-up valuation in your Series B instantly lets your earlier investors revalue their Series A stake upward, on paper, and that markup flows straight into the IRR they wave in front of their next set of LPs.
This creates a powerful incentive that has nothing to do with your long-term health. Your investors need you to keep raising, at higher and higher valuations, on a schedule that fits their fundraising cycle, not yours. Every fresh round at a step-up is a data point that makes their fund look better and helps them close their next one. Whether that round is actually good for you, more dilution, a higher bar you now have to grow into, a valuation that can trap you if the market turns, is a secondary concern.
So the gentle pressure to "go big and raise again" is not only about chasing the power law outlier. It is also about manufacturing the paper markups that fuel the next fund. You can end up raising money you do not need, at a valuation you will spend years trying to justify, primarily because your investors' own fundraising depends on the optics of your last round. A founder who understands this can at least ask the uncomfortable question: am I raising this round because the business needs it, or because my cap table needs the markup?
For decades, the power law bargain made sense to take because there was no alternative. Serious tech companies lost money for years and needed piles of institutional capital just to reach escape velocity. If you wanted to build, you needed the money, and the money came with the power law attached.
Agentic AI broke that assumption. Teams now ship real products, hit meaningful recurring revenue, and reach cash-flow break-even on a friends-and-family round and a couple of strategically lazy founders armed with GPUs and good prompts. The cost to stand up a real software company has dropped by something like an order of magnitude in roughly two years. When you can reach profitability on a few hundred thousand dollars instead of a few million, the entire calculation flips. You no longer have to join the power law risk pool just for a shot at greatness. You can fund your shot yourself.
And once you can fund it yourself, you get to ask a question that used to be unaskable: do I even want a structure built around a distant, improbable, all-or-nothing exit?
This is where the Free Flow LLC comes in, and why it is spreading among AI-native founders. Instead of defaulting to a Delaware C corp engineered for multiple rounds of venture capital and a someday liquidity event, you start as an LLC built around cash generation and tax efficiency first, and VC optionality second.
The alignment is the whole point. In a Free Flow LLC, profits pass through to founders and early investors and are taxed once, on their personal returns, instead of getting hit by corporate tax and then taxed again on distribution. Early losses can flow through to members and offset their other income, which turns the risky experimental years into a present-day benefit for the high-income angels backing you rather than a theoretical write-off trapped at the corporate level. And once the AI engine is throwing off predictable cash, you have real choices. You can reinvest most of it as your own internal VC fund that dilutes no one, while distributing enough to keep founders and early backers financially and emotionally bought in for a longer, more strategic outcome.
Look at what that does to the alignment problem. Nobody at the table needs your 30 million dollar company to gamble itself into oblivion chasing 3 billion. The investors are getting paid along the way. The founders are getting paid along the way. A genuinely good, durable, high-margin business is the goal, not a consolation prize. The Free Flow LLC pays you for building a great company instead of only paying off if you build a unicorn.
None of this means venture capital is bad or that you should never take it. For the right company, a real shot at a nine-figure or larger outcome, a market that rewards moving first and spending hard, a plan that genuinely needs institutional scale within the next year or two, venture capital is the correct tool and the power law works in your favor. The QSBS benefits of a C corp alone can be life-changing if you are truly on a path to a very large exit. And you can always start as a Free Flow LLC and point the nose of the plane toward a C corp later, converting when the capital markets actually demand it and walking into that first priced round with real revenue instead of just a deck and a dream.
The point is simpler than pro-VC or anti-VC. It is this: the power law is the law your investors live by, and it does not automatically align with the law you want to live by. Understand it before you take the check, not after. Know whether you are signing up to be the outlier that returns the fund or just another necessary data point in someone else's portfolio. In a world where AI has made it genuinely cheap to build, you finally have a real choice about which game you play. Make it on purpose.
