Write down who it is for

Try finishing this sentence without hedging: "This is for ______, who is trying to ______, and right now they do it by ______." Most people can start it and cannot finish it. The first blank comes out as a job title, the second turns into a feature list, and the third — the one that matters most — gets skipped entirely, because it requires knowing what your customer's Tuesday afternoon actually looks like before you existed. That unfinished sentence is not a branding exercise you are behind on. It is the reason your ad copy takes four hours to write, your channel picks feel arbitrary, and two reasonable people on the same team can argue about a headline all week without either of them being wrong.

Here is the mechanical version of the problem. Every marketing decision you make is a choice between options, and a choice needs a criterion. Which of these two subject lines? Which of these three channels? Should the landing page lead with the price or the outcome? If you have a specific person in mind, most of these answer themselves in about ten seconds, because one option is obviously written to them and the other is written to nobody. Without that person, every one of these becomes a matter of taste, and taste is slow, inconsistent, and impossible to delegate — to a teammate, a freelancer, or an assistant drafting your first version.

The vagueness is not laziness. It is usually fear of leaving money on the table. Writing "for finance teams at mid-market manufacturers who still close the month in spreadsheets" feels like turning away every other kind of customer, and when you have eleven customers total, turning anyone away sounds insane. So the copy widens until it fits everyone, which means it grips no one, and the wide version is genuinely worse at attracting the very people you were trying not to exclude. You did not open the door to more people. You made the sign on the door harder to read from the street.

a sentence that could describe anyone will not convince anyone.

Try it as arithmetic. Suppose a page gets a thousand visitors a month and converts at one percent — ten signups. Now suppose the narrow version cuts your addressable traffic in half, to five hundred, because half the people who used to arrive now self-select out before clicking. If speaking directly to the remaining half lifts conversion to three percent, you get fifteen signups from half the traffic. These are made-up numbers, not a measurement, and the point is only the shape of the trade: you are not choosing between more and fewer people, you are choosing between a small share of a big room and a large share of a small one. The second one is also cheaper to serve, because everyone who arrives already understands what you are.

Where the answer actually comes from

The good news is you are not supposed to invent this. You are supposed to read it off evidence you already have, and the evidence lives in four cheap places. Your existing customers: not all of them, the five who renewed without being chased or who told someone else about you. The words people type: whatever search terms already bring you clicks are your audience describing their problem in their own vocabulary, unprompted. The pages they land on and stay on — one page out-converting the rest is usually a signal that the page happens to describe one specific situation well. And the direct answer, which is still the most underrated of the four: [ask them how they found you](/blog/ask-them-how-they-found-you), and while you have their attention, what they were trying to fix that day.

Then write the sentence, one paragraph maximum, in plain words, somewhere you will actually see it. Not in a strategy deck. In the document you open when you write anything. It should contain a situation, not an adjective — "runs payroll for forty people with no HR software" beats "growth-minded SMB" every time, because the first one tells you what to write about and the second one only tells you how to feel. If your sentence contains the words "modern", "innovative", or "growing", you have written a mood, not a person.

The payoff shows up in a place people rarely predict: it makes the AI-drafted first version usable. A draft is only as focused as the brief behind it, and "write a launch email for our product" is not a brief — it is a request to guess, and the guess will come back readable, competent, and addressed to no one in particular. Hand over the situation instead, and what comes back is specific enough to argue with, which is the only kind of draft worth editing. The same applies to the human version of this: a freelancer with your sentence writes better copy in an hour than one without it writes in a day.

It also gives you a clean way to be wrong. A vague audience can never be disproven, so it just sits there for two years. A specific one makes a prediction — these people, this problem, these words — and predictions get tested. When the ad promising one thing lands on a page describing another, [the mismatch is usually an audience disagreement](/blog/when-your-ad-and-your-page-disagree) that nobody wrote down, showing up as a conversion rate. So write yours today, date it, and put a note in your calendar for ninety days out to read it again next to your actual customer list. If it still describes the people who are paying you, you got it right. If it does not, you now have something far more valuable than a correct guess: a specific, falsifiable idea about your market that reality has just corrected for free.