This is a lightly edited version of what I sent out to our LPs a few weeks ago in our Q4 '25 management letter. THE POWER LAW PILL Goodhart's Law states that "when a measure becomes a target, it ceases to be a good measure.” Venture capitalists have long discussed the power law of returns, but it finally became canonized when Sebastian Mallaby published The Power Law in 2022, a book analyzing the history and strategy of VC. I would argue that this book, plus @JoshuaKushner's @InvestLikeBest podcast episode in 2023, hallmarked the beginning to what I call the “power-law-pilled” venture capital market. Venture investors haven’t just taken the power law pill that “only a few companies matter” and drive all the returns: they’ve swallowed the whole bottle. During the SaaS era, any company that went $0 to $1M ARR felt like they could raise a Series A, regardless of the quality of the end market. Many companies supposedly “mattered,” a boon for seed managers where downstream capital was available for lots of companies in their portfolio. There are over 700 unicorns from this time period, most of whom will never return capital to their early investors. About a year after the end of the ZIRP era, Kushner went on the ILTB podcast talking about “Fifth Avenue” companies and the idea that quality compounds and what you really want, especially at growth, is the best asset, even if you are paying what seems like a high price. The power law means that all the returns go to the top companies in the market, so those are the only ones you want to be in. Few remember this now, but at the time he said that, many of their large investments - OpenAI at $29B and eventually $80B, Stripe in a down round at $55B, Ramp in a down round at $5.5B, etc. - looked overpriced and were highly contrarian. @ThriveCapital has obviously been proven right in a big way since and the power law, “Fifth Avenue,” and "concentration" are thrown around at every VC associate coffee as standard nomenclature for how to do the job well. Here are some of the changing market dynamics occurring now because of the power-law-pilled orientation of VC firms: - At early-stage, any price is deemed justifiable if the team or market is good enough. Carta data shows upper-quartile valuations at seed rising from $17M to $30M, while the top decile deals have gone from $24M to $50M. The logic goes something like: if the company has the genetics to be a power law company, does it really matter if the seed is at $10M, $30M, or $70M? There’s 100x upside if it truly hits the power law! This is obviously a recipe for potential sloppy thinking amongst investors, but it’s impacting the game on the field today. - Many firms that used to consider the Series A as their bread and butter would now rather wait until the Series B or even Series C. They want to see who has emerged as the “power law winner” if a space still feels early. Conflicting yourself out of the power law company is one of the worst crimes you can commit as a multistage. Marc Andreessen has even argued before that the best strategy for VCs is to be polygamous with their investments in big categories, so you’re starting to see funds loosen their restrictions if they can. - The only exception to the previous rule of waiting is if everyone agrees this is an important category for AI. Categories are either kingmade and important (e.g., ITSM, ERP, RL-as-a-service, code) or completely unimportant, where, regardless of traction, you can’t get a top-30 firm to do your Series A. There’s a reason why Hemant Taneja at General Catalyst will go on a podcast and say T3D2 is dead: what matters is hypergrowth, category importance, funds being able to pile lots of capital into a company, and the consensus belief this company can be a power law winner at scale. Kingmaking and the overfunding of categories with 2-4 legible players is rampant in the market right now. So what does all this mean for seed companies at large? You’re starting to see a lot of this “power-law-pilled” orientation borne out in the data: - There is a ballooning in the top end of price in the seed market, as previously mentioned - The time between seed and Series A is elongating despite the fact companies are growing faster than ever - Fewer seed companies are able to raise Series A’s - Capital is chasing and concentrating in fewer names, not just at the very late stages Companies that are not yet “legible to large pools of capital” (see @WillManidis' excellent piece) must get further with the money they’ve raised and expect tougher fundraises. And as a corollary, companies that are competing in obviously good categories, but against “legible to capital” teams, must realize that they are not on a level playing field. They must have some sort of unique insight that will help them leapfrog their competitors because the other company might raise three fast follow rounds before they are able to raise one. We will see more companies fail, pivot faster, or in the best case scenario, get acquired early, due to the bifurcated funding market. As early-stage investors, we’ve always been laser-focused on team and end market. By picking great teams playing in important end markets, many of our companies already fit what late-stage investors are looking for and currently funding. However, we’ve also had to help some founders navigate an unclear or seemingly unfair funding market that feels different than the one they were told about a few years ago. Markets are interesting and constantly in flux. How the venture market currently views companies is not how they will necessarily view them in 2-3 years. And it's worth noting that some of this correction in funding was long overdue - those 700 unicorns and the idea that almost any company could matter was also clearly wrong. It's also worth asking if the seed fund market of 2006-2021 was an anomaly where history only happened once. To confirm and reaffirm, we ALSO believe in the power law. We just expect it to not always show up so neatly and obviously in the early days (i.e. pre-PMF) than some other market participants would like to believe. Our focus will always be on partnering with the spiky, out-of-distibution founders who build category-defining companies. And as we continue to evaluate the changing market dynamic, we will jump when we see a chance to invest in those opportunities – even when they sometimes happen to be less currently legible than what cookie-cutter Silicon Valley is excited about. We remain highly confident in our ability to pick those special people and markets and expect this to continue to pay dividends over the long run in our returns. -Pratyush