Everyone says they want to invest in the “best founders.” But most have no reliable way of identifying them. [Even Sequoia’s best ever fund had a 50% write-off rate].
Early stage investing requires dealing with so much uncertainty that investors create heuristics to navigate it. Today (for some), it’s investing in AI researchers. A few years ago it was backing alumni of Ramp/Palantir etc. More recently, investors seek out “spiky” founders (or even “broken/traumatized” founders).
These heuristics aren’t useless. But after being an early-stage investor for eight years, I believe these founder frameworks are overly blunt filters. First, they can be gamified by founders during pitches. Second, true founder quality can be difficult to glean in just a few meetings, even if you are a world class investor.
But most importantly I believe that an over reliance on founder quality (which is generally hard to measure) means we under weight the factor that has even more impact on long-term success: whether your business is predicated on an secret, or insight, about the world.
Secrets
Peter Thiel memorialized the idea of secrets in Zero to One. It’s a captivating concept: that the best companies are built on an insight that the world disagrees with them on.
I always found it to be an intellectually interesting theory, but with experience came to realize it was also true in practice. Most of the successful companies in our portfolio were built on contrarian insights. A few quick examples:
- PostHog: when PostHog launched, product analytics was already a crowded category (think GA, Mixpanel, Heap, Amplitude etc). The founders’ insight was that competing in crowded markets is actually a good thing to do, because it definitively means there is customer demand. (Obviously, this needs to be paired with some other advantage to cut through the noise). In PostHog’s case, their first advantage was their ability to build product. They quickly went multi-product (e.g. AB testing, session replay, feature flags) and positioned themselves as an all-in-one alternative to various point solutions (which allowed them to undercut competitors, because they could make up margin across other products). Beyond product, they had superior customer acquisition to their peers, acquiring hundreds of YC companies each year at inception and growing with them forever
- Padsplit: Padsplit’s marketplace allows real estate investors to invest in low-income housing, turning a two bedroom house into three bedrooms which can be rented out individually. Most investors (incorrectly) thought that low-income housing was a worse asset class than short-term Airbnb rentals. But Padsplit’s founder, a seasoned real estate investor, realized the opposite was true: despite the tenant demographics, Padsplit properties could generate superior yields due to there being more rooms to monetize as well as higher year-round occupancy. Moreover, the business benefits from a network effect at scale, as more investors attracts more tenants and vice-versa
- Statusphere: influencer marketing has been talked about as an emerging marketing category for a decade. But most entrepreneurs focused on pairing brands with hugely famous influencers who could shill products to a wide audience. Statusphere’s founder, a marketer by trade, realized that the opposite was true. In the age of TikTok - where going viral was about engagement rather than follower count - micro influencers (those with smaller, but loyal followings) were more performant than macros. As a result, she steadfastly built a network of 100K+ micros and is now the preferred partner for giant consumer brands who can’t achieved scaled influencer campaigns elsewhere
These companies are all led by fantastic founders, but there wasn’t a clear heuristic you could use to pick them all. And in each case, investors were dismissive of the founders’ insights for years, until their ideas were proven true and became mainstream. Secrets are in shorter supply than fantastic founders.
Having a Why Now is not the same as a secret
AI provides an incredible Why Now for many businesses - namely, that there is technology available that allows you to solve a problem in a meaningfully better way than how things used to be done. There are clear 10x improvements to be had over the status quo. But a Why Now isn’t a secret if it is widely believed. It’s why you find 50 companies going after every plausible-sounding idea (e.g. AI for law firms, AI customer service).
Even if growth is spectacular for these businesses in the short term, there are big question marks over their long-term durability - and long-term is where most of the value gets built. [I think this is also why kingmaking is so prevalent today: to dissuade competition in markets that would otherwise trend towards perfect competition]
Where can founders look for secrets?
So where can founders look for secrets? It’s not easy to find a secret because it requires uncovering a truth about the world that isn’t widely known (which I think explains why secrets are in shorter supply than great founders). Nonetheless, I think there are a few ways to find a secret today:
- Industry expertise: Being an industry expert has never been more valuable. Deep industry expertise (coupled with AI-first engineering) gives you the knowledge of exactly what to rebuild (with AI). It also gives you a product roadmap that can’t be easily replicated. Many AI-first systems of record could be rebuilt with insights from industry insiders
- Market analysis: Study a market where there is consensus that it is too small, weird or ugly. Question every assumption deeply and see whether there is an unlock with AI
- E.g. selling to SMBs with low historical willingness to pay. Perhaps there is now an opportunity to tap into labor budgets, flipping an uneconomical segment into a fruitful one
- E.g. it would’ve been easy to dismiss the idea of selling data to the big labs as a weird market with a concentrated buyer base. But that would have hugely undervalued the labs’ insatiable demand for data to improve their models
- Incumbent analysis: Study an industry where there is consensus that the incumbents are deeply entrenched. Question every assumption to see if they are still true. Is there a better way to do things? And is there an innovator’s dilemma preventing incumbents from changing course?
- E.g. do you really need a real estate agent to help you buy or sell your house, or could AI now market your home, price it accurately and also assess the offers for you?
If the above resonates and you think you’ve uncovered a secret about your industry, I’d love to meet and learn more. Feel free to reach out to me at samit@1984.vc