every first-time founder thinks the product is the moat. build the best product, attract users, win the market. the market for talent, funding, and press reinforces this — people love technical stories about novel technology beating incumbents. in reality, outside of a narrow set of technical-deep-tech categories, the best product almost never wins. the company with the best distribution wins, and the best product becomes a secondary factor that either amplifies the distribution advantage or fails to overcome it.
the cleanest proof of this is how many category leaders are the second, third, or fifth technically-best products in their space. google was not the first search engine. facebook was not the first social network. instagram was not the best photo app. zoom was not the best video call tool in 2019 — it was the one every it department already had deployed when march 2020 hit. in every one of these cases, the winner arrived with a distribution advantage and took over with a product that was fine-but-not-exceptional. meanwhile, the technically superior competitor lives on as a footnote.
if you look at the canonical tech-history essays about why this happens, the answer is always some version of "users don't evaluate products the way critics do." they evaluate products in the narrow context of what they already use. the product with a distribution advantage has an unfair edge: it gets tried by more people, collects more feedback, iterates faster, and the users who try it don't have to quit anything to use it. the technically superior product, sitting at a site that requires you to find it, sign up, and swap, has to overcome every piece of friction in the user's workflow to compete. the gap in friction is usually much bigger than the gap in product quality.
the practical implication, which took me a long time to internalize. when you're picking what to build, pick the thing where you have a distribution edge, not the thing where you have the strongest product hypothesis. a mediocre product with a channel is worth more than a brilliant product without one. a brilliant product with a channel is worth an order of magnitude more. the reason y combinator-style advice is "build something your friends want" is that it's a way of putting the founder into the channel — they already know how to reach the early users, because those users are their friends. this works. "build the thing with the biggest tam" doesn't, mostly because the founder has no idea how to reach anyone in the tam.
the other practical point is that distribution is durable in a way product is not. a product advantage can be copied in six months. a distribution advantage — a list of customers who already trust you, a landing page that ranks for a high-intent query, a twitter following, a deployment base inside enterprise — takes years to build and years more for a competitor to replicate, if they can at all. most moats in real companies are distribution moats dressed up as product moats. salesforce isn't winning because their crm is better than the alternatives; it's winning because the average enterprise customer has half a dozen integrations on top of salesforce that it would be painful to rewrite. the ui is the last thing anyone switches away from.
this has implications for startup strategy that founders tend to resist. one: spend a lot of time thinking about how you're going to reach users before you write the first line of code. this is boring, it feels like the wrong thing for engineers to do, and it's almost always the single highest-return hour of early-startup planning. two: build features that compound your distribution, even if they aren't the "most valuable" features from a pure-product perspective. referral loops, shareable artifacts, public-facing surface areas — these are often a better use of the fifteenth engineer than the next line item on the product backlog. three: when you consider a pivot, weight the distribution you'd be throwing away. most pivots are worse than the founders think because they lose the channel along with the product.
the remaining 20% of the moat, in my rough accounting, is product quality and operational excellence. they matter. they matter especially in categories where users cost a lot to acquire and stay a long time, because you can only afford to acquire an expensive user if you can retain them, and retention is a product problem. but in practice, the founders who think about product 90% of the time and distribution 10% of the time end up building excellent products that nobody ever sees. the founders who invert the ratio — distribution 70%, product 30% — usually build something great, because a product people are using gets better faster than one people aren't.
if you want one contrarian take to leave this post with: the question "is this the best product" is less important than the question "do i have an unfair way to get this into someone's hands." if the answer to the second is yes, the first sorts itself out. if the answer to the second is no, the first almost never saves you.