Target Market

Ask a product team who something is for and for a long time the answer was anyone who needs it. That answer isn't wrong, and it's also useless. Aim at everyone and nothing is left to decide what gets built next, which request gets picked up first, or whose vocabulary the ad copy should borrow.

Pagedeck's early backlog meetings looked exactly like that. On one side sat a request from a large manufacturer: put an approval chain on every document and keep a record of who signed off. On the other sat a request from a three-person design studio: let clients view a document from a link without signing in. Both came from paying customers, both were reasonable, and both would take about the same time to build. When Mia asked which one went first, the only arguments in the room were the size of the company that asked and the volume of the person relaying it.

Repeat that meeting enough times and the product starts to resemble both requests a little. It has approval chains and it has link sharing, and it does neither one well. The manufacturer says the sign-off record isn't detailed enough to survive an audit, and the studio says the screen is busy enough that clients get lost. The same problem spreads outside the product. Ravi has to fit security and simplicity into one line of ad copy, and Sofia walks into a meeting without knowing which of the two stories the person across the table wants to hear.

A target market is what you settle in advance so that meeting has an answer. You commit to solving one group's problems for a while, and everything outside it gets heard politely and ranked lower. If product strategy is the work of choosing what to do and what not to do, the target market is the floor those choices stand on. A product vision can state the long-term destination in one sentence, but with the audience left blank, strategy collapses back into a contest of preferences.

Faster decisions aren't the only return. Two more things become possible. First, the comparison set gets specific. Competitive analysis lines up rival products by feature and price to find the gaps, and which rivals go on that list depends on which market you claim to be in. Second, success becomes something you can rule on. The state of a built product matching demand is called product-market fit (PMF), and the phrase carries the condition of a particular market inside it. Name no market and there's no way to say whether anything has met it. A retention number of 40 percent is neither good nor bad on its own; it only means something once you can say which group it describes and whether that group is the one you chose.

Four terms sit underneath, in two groups. Three of them decide who the group is; one measures how large that group is. Deciding comes first and measuring second, because there's nothing to measure until there's an audience.

Deciding who runs in three steps.

1. Segmentation — cutting what looked like one market into clusters that behave alike. There are many knives: company size, industry, the tools already in use, the way documents get handled. Within the same collaborative-document market, a 20-person company and a 5,000-person company want close to opposite things. The first wants almost nothing to configure; the second wants permissions sliced finely. Without the cut, the two collapse into one average. A survey result of 3.2 out of 5 on configuration describes neither group. Cut it and one side reads 4.1 and the other 2.3, and only then is there a question worth acting on: which of the two this product intends to serve.
2. Ideal customer profile (ICP) — the written conditions for the customer who fits best and stays longest, picked out of those clusters. It's written at the level of the company rather than the person. Something like: 30 to 200 employees, builds software, already runs at least one collaboration tool, has someone whose job includes documentation. With that on paper, an inbound inquiry can be judged on the spot. The conditions worth writing are the ones that split cleanly rather than the ones that sound appealing. "A company that writes a lot of documents" can't be checked; "a company already running an internal wiki" can.
3. Beachhead market — the single square you set out to take first. Even when several clusters fit the profile, splitting effort across them means finishing first nowhere, so you pick the one you can win soonest and fill it. The name comes from securing one landing point and widening outward from there. Winning here means everyone in that group knows the product, and that reputation is the force that carries into the next square. The line for having won gets set in advance too — something like a share of spend inside the group crossing a stated percentage, or a majority of new inquiries arriving through people already inside it.

Measuring size is one term. TAM/SAM/SOM looks at the same market in three shells. Everyone who could need the product is the outer shell, the portion the current product and sales channels can actually reach is the middle, and the slice realistically winnable over the next few years is the inner one. The split into three exists because the outer number alone decides nothing. A three-trillion-won market tells a team nothing about what to do next quarter. The further in you go the smaller the number gets and the more useful it becomes. Even in a market that large, a product that runs only in a browser and takes only domestic payments reaches a fraction of it. Narrow again to the companies this year's sales headcount can actually meet and what's left is a number worth aiming at. Investors ask for the outer number because the market has to be large for the company to grow into it; working meetings ask for the inner one because only that becomes a plan.

Both groups have to be present because each blocks a different failure. Decide who without measuring and you can land somewhere so small that the product is loved while the company stalls. Measure without deciding who and you pick a market with an impressive number where nobody inside it feels the product was made for them. The beachhead Pagedeck chose was software companies under 50 people, and the winnable slice inside it came to 3,400 companies domestically. Not a large number, but the math said 400 of them would fund the next stage, and that math set the feature order and the hiring plan together. None of this happens in one sitting. In practice it runs over weeks — cut, choose, measure, pick one square — and when the measurement comes back too small, the chosen conditions widen and the loop starts again. The loop is worth running on a fixed schedule rather than once. A market that fit two years ago can drift as the product grows, and the group that was hardest to win early often becomes the obvious one later.

"This request is outside our target market. Log it, but it's not going on the roadmap."

The request may be reasonable, but it isn't the problem of the group currently being aimed at, so it drops in the order.

"If the target market is every team that collaborates, nothing is decidable. Let's narrow it to software companies under 50 people."

An audience that broad can't function as a criterion, so the proposal is to cut it down to something judgments can be made against.

"I hear the whole market is three trillion. What can we actually reach this year?"

A request for the inner number instead of the outer one.

Pagedeck narrowed its target market for the first time right after its Series A. Until then the customer list mixed ad agencies, hospitals, schools and software companies. Grace called the meeting and asked not where the team wanted to go but where it was already winning. The numbers Sofia brought answered it. Software companies were still around a year after signing 81 percent of the time; everything else came in under half. Most inbound referrals traced back to that same group as well. Nobody in the room had picked that market on purpose; it had simply been the one where the product kept working, and the meeting turned an accident into a stated choice.

Three things changed together once the decision landed. Mia dropped the approval-chain work aimed at large manufacturers and pulled forward the work connecting documents to the tools engineers already use. Ravi took the word security out of the ad copy and put in the line about code and documents living on one screen. Sofia started checking headcount and industry before anything else, pointing companies outside the conditions to the free plan instead of booking a meeting. What made the meeting worth holding was that those three decisions came out of one sentence rather than three separate judgments.

The most common mistake is reading a narrowed market as revenue thrown away. It sounds like refusing to sell to nine of the ten groups you could serve, which is a natural read and a strong one when money is tight. It works the other way around. Narrowing doesn't turn away inbound interest; it decides what gets built first and where people get assigned. The broad team finishes second or third in all ten groups, while the narrow team finishes first in one and uses that standing to enter the next. Leave the mistake in place and the product accumulates features while no group calls it their tool. In that state there's no argument for raising the price and no reason for referrals to appear. What a broad team actually experiences is less revenue growth than scattering: ten groups' requests land in one backlog, engineering ships unrelated things every quarter, and none of them fully satisfies anyone before the next quarter starts.

The second is treating the current customer list as the target market. These are the people actually paying, so no evidence looks firmer. But an early customer list is mostly the product of a founder's network and whichever channels happened to connect. Six ad-agency customers often trace back to the founder having worked at an ad agency. Take that list as the audience and an accident gets promoted into a strategy. Use the list as material and make the judgment separately: which group stays longest, which group generates referrals, which group pays full price. It isn't unusual for the answer to push half the current list outside the target. That doesn't mean showing those customers the door. It means the next round of building isn't aimed at them. Money already taken carries the promises that came with it, so what they use keeps working; what changes is the direction of new effort. Without that distinction stated plainly, a decision to narrow gets relayed as a decision to drop customers, and once that version spreads, prospects outside the list hesitate too. Saying it out loud in the customer's own terms costs one email and prevents a quarter of guessing.