Audience
How to define your target audience in four steps
Defining a target audience means narrowing a market down until one group is left that shares the same problem, can be reached through the same channel and judges an offer by the same yardstick. A definition is usable when it passes five criteria: it is measurable, big enough, reachable, distinct from the rest of the market and workable with the time you have. Start from the problem, draft three candidate groups, run them through those five criteria, pick one, and then correct the description with ten conversations. "Everyone who owns a bike" is a market. "Commuters who ride daily and cannot afford a breakdown" is an audience.
"Everyone" is the most expensive answer
Asked who the customer is, most founders answer with a market. The answer feels safe, because it excludes nobody. It is the expensive one, because every following decision loses its ground: which price is right, which channel is worth the money, which sentence goes on the first page. All three depend on the group, and if the group is everyone, all three become guesswork.
There is a second, quieter cost. A message written for everybody has to stay general to stay true, and a general message is one nobody recognises themselves in. Narrowing therefore does not reduce your chances, it is what makes a clear sentence possible in the first place. The point is not to serve fewer people, it is to be understood by the first ones.
Three definitions on the test board
Everyone who owns a bike
That is a market, not a segment. You can count it but you cannot serve it: no channel reaches these people together, and no sentence is true for all of them.
Portland commuters who ride daily and cannot afford their bike to be out of action
Workable. The group is countable, large enough for a workshop, reachable along commuting routes and through employers, and it judges by a yardstick of its own: downtime.
Carbon-frame road cyclists in one neighbourhood who train on Tuesdays
Too narrow. Reachable and clearly bounded, but the size can neither be estimated nor support a business. A target audience may be small, it just has to stay countable.
The bar widths illustrate how the group shrinks, they are not market data. What the board is for is the middle row: a definition passes only when all five criteria hold at once.
The five criteria, in plain words
The set goes back to Philip Kotler and has outlived every marketing fashion since, because each criterion rules out a specific way of fooling yourself.
1Measurable
You can put a number on the size, even a rough one, and say where it comes from.
Rules out groups that only exist as a feeling.
2Big enough
Serving the group covers your costs, at a price this group would pay.
Rules out audiences that are precise and unprofitable.
3Reachable
You can name a place, a channel or an occasion where these people can be addressed.
Rules out groups you can describe but never meet.
4Distinct
The group responds differently from the rest of the market to price, wording and offer.
Rules out a split that changes nothing about what you do.
5Workable
With your time and money you can actually serve this group, starting now.
Rules out a definition that only works with a team you do not have.
Four kinds of attribute, in ascending usefulness
Most descriptions use only the first two, because they are the easiest to get. The two below them are the ones that explain a purchase. Use all four, but let the last one decide.
1Geographic
Where are these people?Examples
Country, region, city, neighbourhood, distance from your base
For anything delivered in person this is the hardest boundary there is. It decides travel time, ad budget and how many jobs fit into one day. For digital offers it is usually the weakest of the four.
Usually drawn too generously. Serve a whole city and you spend the day in transit.
2Demographic
Who are they, from the outside?Examples
Age, occupation, income, household size, education
These are the easiest attributes to obtain and the easiest to overrate. They help with whether someone can afford the price, and with setting up ads. They say little about why anyone buys.
Two people with identical profiles can buy for completely different reasons.
3Psychographic
What matters to them?Examples
Values, priorities, risk appetite, self-image, daily rhythm
This is where the reason sits for choosing one offer over another. These attributes do not come from statistics, they come from conversations. They are the raw material for your wording: they decide which sentence gets heard at all.
Do not invent them. What goes here should trace back to something a real person said.
4Behavioural
What do they already do today?Examples
Frequency, trigger, current workaround, willingness to pay, reasons to switch
The most telling of the four, because it is observed rather than asserted. Anyone already spending money or time on the problem has proved it exists. Those people are your first customers.
People who do nothing today rarely feel the problem urgently. Treat that as an exclusion, not a challenge.
Selling to companies rather than people, the first two are replaced by firmographics: industry, company size, revenue, number of sites, and above all who decides and who signs. The other two still apply, they just belong to a role instead of a private person.
In four steps to a usable definition
The first three fit into one afternoon. The fourth takes a fortnight and is the only one that turns the description from a guess into something you know.
Start from the problem, not from the person
30 minutesWrite the problem down in one sentence first, with no description of any person in it. Then ask: for whom is exactly this problem expensive, urgent or embarrassing? Go the other way round, starting with a picture of a person, and you will look for a matching problem afterwards and always find one.
Result: one problem sentence and a first list of groups it hurts most.
Draft three groups and set them against each other
45 minutesDescribe three possible groups concretely enough that you can say where you would meet each of them. Use all four types of attribute, and the behavioural one at the very least. Three candidates beat one, because they can be compared and force you to name the differences.
Result: three descriptions of two to three lines each, written side by side.
Run them through the test board and pick one
45 minutesTake each group through the five criteria. If a group fails on reachable or workable, drop it immediately, however good it feels. At the end you pick exactly one group for the coming months. The others are not discarded, they wait in writing for their turn.
Result: one chosen group, two parked, each with a one-sentence reason.
Hold ten conversations and correct the description
1 to 2 weeksFind ten people from the chosen group and ask about their daily life, not about your idea. Then rewrite the description. Something almost always shifts: the trigger turns out to be a different one, or the real pain sits one level deeper. That correction is the point of the step, not its failure.
Result: a revised description in which every sentence traces back to a conversation.
Market, segment, target audience, persona, ICP
The five terms get used as if they meant the same thing. They describe five different levels of detail, and only one of them is a decision.
| Term | Question it answers | What it covers | What it is good for |
|---|---|---|---|
| Market | Who could conceivably need this kind of thing? | Every possible buyer of a type of solution | An order of magnitude for market sizing. Not an instruction for anything. |
| Segment | Which parts of the market behave differently? | Several separable groups inside the market | A map of the market. It still does not say where you start. |
| Target audience | Which segment do you serve first? | One segment, deliberately chosen | A decision. It sets the offer, the price, the channel and the wording. |
| Persona | What does a typical person in that group look like? | One invented individual, assembled from real conversations | An aid for copy and design. Not a substitute for the group. |
| Ideal customer profile | Which company fits us best? | B2B only: the attributes of the best customer company | A selection rule for sales. It describes companies, not people. |
When the description becomes advertising
Describing a group for yourself is free of rules. Turning that description into filters on an advertising platform brings two limits with it. The first: health, religion, political opinions, trade union membership, ethnic origin and sexual orientation count as special categories of personal data in the EU. Processing them is prohibited apart from narrow exceptions, and building an ad audience on them is not one of the exceptions.
The second concerns minors. Online platforms in the EU may not show advertising based on profiling when they can reasonably assume the recipient is under 18. If your audience includes young people, plan for reach through context instead: the place, the occasion, the school, the club, the topic of the page. That works without personal profiles and stays available.
General orientation, not legal advice. The rules differ by country and change; check what applies where you advertise.
Five mistakes that keep the definition useless
Each of them produces a description that reads well. None of them produces one you could act on.
Deriving the audience from the solution. Build first and then ask who might need it, and you find a group that fits the solution instead of a solution that fits a group. In order: the problem comes first, the group second, the solution last.
Excluding nobody, for fear of losing someone. A target audience that excludes nobody reaches nobody. The sentence "this is not for you" is uncomfortable and is exactly what sharpens the message for everyone else.
Describing a persona instead of a group. An embellished profile with a photo, hobbies and a favourite show feels like work and is not. A group only becomes testable once you can say how big it is and where you meet it.
Never touching the definition again. It is an assumption with a date on it, not a resolution. If your description after fifty sales conversations reads exactly as it did on day one, you either hit the mark or did not listen, and the second is more likely.
Confusing ability to pay with willingness to pay. That a group could afford your offer says nothing about whether it wants to spend money on it. The evidence sits in today's behaviour: what does this group already pay for that comes close to the problem?
Frequently asked questions
How do I define my target audience?
In four steps: write the problem down in one sentence, with no person in it. Draft three possible groups that feel this problem most. Test each group against five criteria (measurable, big enough, reachable, distinct, workable) and pick exactly one. Then hold ten conversations with people from that group and correct the description. The fourth step matters most: before those conversations, every audience description is a guess.
How narrow should a target audience be?
Narrow enough that one sentence is true for everyone in it, and wide enough that the size can still be estimated and a business can live off it. If estimating is hard because there is no number and no list to work from, the group is cut too tight. Early on, too narrow is still the cheaper mistake: a narrow group can be widened, a wide one costs months before the error shows.
What is the difference between a target audience and a persona?
The target audience is a real set of people with shared attributes that can be counted and reached. The persona is an invented individual, assembled from real conversations, so that you have somebody in mind while writing. The persona is an aid and does not replace the group: you cannot target an ad at a persona, you can target one at a group.
Can I have more than one target audience?
In the long run yes, at the start rarely worth it. Every additional group means its own wording, its own channels and often a slightly different offer, all within the same amount of time. There is one exception: when the user and the payer are different people, an employee and a purchasing department for instance, both belong in the description from day one, because both have to say yes.
Where do I get data about my audience with no budget?
From four sources that cost nothing. First, conversations with ten people from the group, the richest and the most uncomfortable source. Second, reviews and forums where those same people write about existing offers. Third, public statistics for the order of magnitude, from national statistics offices or industry associations. Fourth, the audience size an ad platform reports for a given set of filters, as a rough indication and not as evidence.
Read next
Is your audience carrying the idea?
The MARKET SCOUT sizes the segment you named, the STRATEGIST checks whether you can reach it. 4 to 12 AI experts, in minutes, from € 5.99.
Assess my idea →