Choosing which market to enter when you can afford one

A business with limited money and one team can enter one market properly or several badly. The temptation is to keep options open — a bit of effort in three cities, two customer segments, both a consumer and a business version — on the reasoning that spreading effort spreads risk. In practice it spreads effort below the level at which anything works. Every market needs enough presence to be noticed, enough learning to get the offer right, and enough patience to survive the period before word of mouth starts. Split three ways, none of them gets any of that.

So the decision is worth making deliberately, and worth writing down. The value of a written framework is not that it computes the answer; it is that it makes the reasoning inspectable later, when the market you chose is harder than expected and you need to know whether the logic was wrong or the execution was. A decision made on instinct leaves nothing to review, which means the next decision starts from scratch.

Market size is the obvious one and the least discriminating, because a big market is only relevant if you can reach a slice of it. Pair it with the cost of reaching that market: is there a channel — a location, a platform, a community, an existing relationship — through which you can put your offer in front of these buyers at a cost you can sustain? A small market with a cheap, direct channel beats a large one you can only reach by outspending incumbents.

Competitive intensity is third: how many businesses already serve these buyers, and how well. Fourth is regulatory and structural friction — licences, certifications, compliance obligations, minimum capital, distribution rules — which is not a reason to avoid a market but is a real cost and a real delay that has to be planned rather than discovered. Fifth, and most often skipped, is willingness to switch. Buyers who are content with their current solution are expensive to move regardless of how good you are, while buyers who are actively dissatisfied or currently unserved are cheap by comparison.

Put your candidate markets across the top of a table and the five criteria down the side. Score each cell out of five, and — this is the part that does the work — write one line of evidence next to each score. "Reach: 4, because there is an established trade association with a member directory and a paid listing" is a claim someone can challenge. "Reach: 4" is a feeling with a number attached. If you cannot write the evidence line, the honest score is a question mark, and a column full of question marks is itself a finding about how little you know.

Weight the criteria if some genuinely matter more to you, but resist elaborate weighting schemes. The arithmetic is not where the insight is, and a framework precise enough to look authoritative invites more confidence than the inputs deserve. What the table reliably does is expose disagreement inside a team, surface the criterion where a favoured market scores badly, and stop the loudest preference from winning unexamined. If the totals come out nearly level, that is genuine information too: pick on a different basis, such as which one you can exit most cheaply.

Large markets attract competent competitors, and they attract them early. The buyers in them have been marketed to extensively, which raises the cost of getting attention and lowers the impact of anything you say. Channels into them are priced by what the best-funded bidder can afford. And the sheer variety of buyers in a big market makes it hard to learn anything specific: feedback arrives from many kinds of customer with conflicting needs, and averaging it produces an offer that fits nobody in particular.

A narrower market inverts each of those. Fewer competitors take it seriously, the buyers are less saturated, the channels are often cheap or free because they are communities rather than advertising inventory, and — the underrated part — the feedback is coherent. Serving one specific kind of buyer teaches you exactly what that buyer needs, which is how an offer becomes genuinely good rather than broadly acceptable. Being known within a small market is achievable; being noticed in a large one may not be. The larger market does not disappear while you do this, and entering it later with a proven offer and a reference base is a different proposition than entering it first with neither.

A scoring table is built from beliefs, and some of them are checkable more cheaply than a full entry. Before committing inventory, a lease, a hire or a year, identify the one or two scores the decision actually hinges on and find the cheapest thing that would test them. If the case rests on buyers being dissatisfied with the incumbent, go and ask twenty of them what annoys them. If it rests on a channel being affordable, run a small paid test or ask the community's organiser what a listing costs. If it rests on the absence of a regulatory barrier, read the actual rule rather than asking someone who thinks they remember it.

The useful discipline is to name, in advance, what result would change your mind. A test you would ignore if it came back badly is not a test, it is reassurance, and it is worth neither the time nor the money. Set the threshold before you look. It is also worth agreeing in advance what evidence would make you stop after entry — a volume you have not reached by a certain date, a cost per customer above a stated figure — because deciding that while losing money is a decision made under exactly the wrong conditions.

The scores are judgements, and dressing them as arithmetic does not convert them into measurements. Two people scoring the same market on the same evidence will differ, and the total is only as good as the least reliable cell in the column. Market size figures in particular deserve suspicion: any number describing a market's total value came from somewhere, and unless you know whose research it is, what year it covers, what geography it includes and what it counted, it is not a fact you can plan against. A bottom-up estimate built from buyers you can name and count is less impressive and far more useful.

There is also a limit no framework crosses: it cannot tell you whether buyers will actually buy. That is discoverable only by offering something and watching. What the exercise legitimately delivers is a shortlist, a record of why one candidate was preferred, and an explicit list of the assumptions the choice depends on. Those assumptions are the thing to revisit when reality disagrees — which is a considerably better position than having a strong opinion and no idea which part of it was wrong.