# Where to compete *Know one market better than anyone* --- The better product loses more often than anyone likes to admit. I've watched it from both sides: the challenger with the cleaner, faster, cheaper system that can't dislodge an incumbent whose software looks ten years old, and the incumbent with the ageing product that keeps renewing anyway. The explanation usually offered is inertia, or relationships, or risk aversion in the buyer, and sometimes that's what it is. More often the challenger has misunderstood what they're competing against, and the mistake starts much earlier than the sales process, with where they chose to compete. ## How narrow to go A generalist looks at financial services, or healthcare, or logistics, and builds something broad enough to sell across all of it, which leaves it fitting none of it especially well. The specialist goes the other way, into the corner where the rules are strangest and the generic product fits worst. Jack Henry serves community banks rather than banks: the institutions big enough to need real technology but too small to build it themselves, with their own regulators, their own board reports, their own rules about what a branch manager can approve. Veeva built its software for life sciences under regulation and nothing else. These are sub-segments most people wouldn't think big enough to build a company on, and that's exactly why they hold: too small and too particular for a big generalist to serve properly, deep enough to keep one specialist busy for decades. How narrow is narrow enough? Scale is a frame-of-reference question. From outside an industry you can't tell whether you're looking at one global market, two hundred country-level ones, or two thousand regional ones; from inside you know, or you should, because the dynamics of the customer base tell you. Every market has a minimum scale you need to operate in it properly, and it's set by what serving that customer actually takes, not by how big the number on the market map is. That's why a lane that looks absurdly small from outside can be a perfectly good business, and why you can't enter a big market halfway. Stay in one lane long enough and you get an intimacy with the customer that nothing in a demo can match. The workflows end up carrying the way the customer actually works, configured a piece at a time until the system knows things the customer would struggle to write down themselves. You become the name people in the niche pass to each other, while the wider market has never heard of you. The switching cost that gets called a moat is mostly just this: the customer has built the way they work around you, a piece at a time, and you couldn't have designed it in. ## What intimacy buys you Without it, it's easy to end up solving a narrow problem on the wrong dimension. You look at the incumbent and see that it's clunky, too many clicks, could be sped up - and all of that can be objectively true. But the embedded product usually speaks to customer systems, supplier systems or a regulator in some meaningful way, and an incremental improvement on one dimension doesn't stack up against the risk of disrupting that. It won't typically get voiced, either, because the person facing the clunkiness - the user, the administrator - isn't the person weighing the risk; the CFO is, and the CFO cares about the financial implication and the compliance. Really understanding what causes sleeplessness in your customers gives you a much better chance of innovating on the dimensions where they care. That's the test I'd put on any product plan aimed at somebody else's market: which dimension is this better on, and is it a dimension anyone with sign-off authority actually worries about? ## Enter at the edge Suppose the lane you want is currently held. Attacking the incumbent's centre is a fair fight at best: their product fits, their relationships are deep, their proof points are overwhelming, and you're asking a buyer to leave a product designed for them and take a risk on one that wasn't. Every incumbent has an edge, though: the customers they've accumulated who are slightly wrong for the way they sell, slightly underserved by the product, and not really anybody's priority. A product built for large enterprises gets sold down-market, and the smaller customers get a system heavier than they need and a roadmap aimed at problems they don't have. They're adequately served rather than delighted, and their switching costs, while real, are lower than the core customers', because they've integrated less and customised less. These pockets usually show up in the data before you see them in the market. Segment your competitive win rate by customer profile and most segments cluster around the average, but one sometimes jumps to a multiple of it. Treat that carefully - a handful of wins in a narrow segment is a thin sample and the range around it is wide - but a gap that big usually has a structural cause, and a structural cause is something you can go and find. When you show up there with a product that fits their actual scale and a sales process that treats their deal size as worth winning, the fight moves onto your terms. Win the pocket decisively, let the rest go quiet for a while, and each win makes the next cheaper: the case studies, the references, the objections your team has already heard. The edge expands into the segments next door, and what was the incumbent's territory starts to feel contested. ## What to do once you hold it Say you dominate a niche: 65% share, margins above 25%, returns on capital around 40%. The obvious growth story is the adjacent bigger market, fifty times the size, where a sliver of share would triple your revenue, and the business case writes itself. But you'd be one of forty competitors instead of one of ten, your near-total brand recognition goes to zero at the boundary, acquisition costs multiply, and you'd be up against incumbents who know that market the way you know yours - so the returns that funded everything halve on the way in, and the cash the core generates gets absorbed building a position you may never hold. The less obvious alternative is to grow the market rather than your share of it. Growth comes from new applications for the same capability, from new geographies where regulation is catching up and creating demand that didn't exist five years ago, from adjacent problems your existing customers already ask you to solve. A £20m niche that becomes £50m over a decade, with your share intact, more than doubles you without ever leaving the territory you understand best - and your returns fund that expansion in a way competitors at a twentieth of your share can't match. The same logic works service by service as well as market by market. [[Cintas]] started with uniform rental and now sells mats, first aid, restroom supplies and fire protection down the same weekly route: each service sold to a customer already on the round, each one making the visit more valuable and the relationship harder to leave. Growing the lane can mean deepening what you sell to the customers you already hold as much as widening who you sell to. ## What going narrow costs Going narrow costs you twice. What you've learned about one vertical is close to useless in the next, so you grow by going deeper where you are, or you start the learning over somewhere new, and starting over is a far harder game than the margins in your first market make it look. And the intimacy has a blind side. When customers stay because they're embedded, a product can slide behind the market for years while the retention numbers still read as loyalty. I've taken a book of loyal customers as proof the product was fine and left it a year longer than I should have before admitting it had fallen behind. The stickiness that protects you is also what lets you kid yourself. ---