Overview
“Marketing teams who want to grow” is not an audience. It is a market.
A market is a state people are in. An audience is one person at the moment something happened to them. Copy written to a state has to stay general enough to cover everyone in it, which is why it ends up saying nothing. Copy written to a trigger can be specific, because it only has to be true for the person who just had that thing happen.
This skill produces an audience card. Nothing gets written until every slot is filled with a sentence that could be wrong.
The audience card
All six slots are REQUIRED. An empty slot means the brief isn’t ready, not that the slot doesn’t apply.
1. The person. Their role, and what their week actually looks like at the moment they’d encounter this. Not a title in isolation. A title with a workload.
2. The trigger. The event that makes them look. An event has a date. “Wants to improve their SEO” is a state and fails. “Their agency contract just ended”, “traffic dropped 30% after a core update”, “they got asked for numbers in a board meeting” are events and pass.
3. The current fix. What they do today instead. A spreadsheet, a freelancer, an intern, a competitor’s product, ignoring it.
4. The belief to change. One sentence they currently do not accept, and would have to, before any of this matters. If you can’t state it as a sentence they’d argue with, you don’t have one.
5. The cost of being wrong. What happens to them personally if they choose badly. This sizes everything downstream: how much proof they need, how long they’ll read, how many people they’ll have to convince.
6. Who this is not for. Named explicitly, as a group who might plausibly otherwise think it was for them. “Not for solo founders under $10k MRR. The setup cost doesn’t pay back.”
The two tests
The nothing test. If slot 3 comes back as “nothing, they just live with it”, stop. Either they don’t have the problem, or the trigger in slot 2 isn’t the real one. People with real triggers always have a current fix, even a bad one. Go find it before continuing.
The split test. If any slot has two answers, this is two briefs. Split it.
Averaging two audiences produces copy that is slightly wrong for both. This is the most common way a brief fails, and it stays invisible until launch. Two sharp briefs beat one broad one even when you only have budget to run one, because you can pick.
Output contract
Return the card as six labeled slots, then one line: “Write for [person] at the moment [trigger]. Change their mind about [belief]. Not for [excluded].”
That sentence goes at the top of the brief. Everything written afterwards gets checked against it.
Anti-patterns
| Pattern | Why it fails |
|---|---|
| Demographic personas (“Marketing Mary, 34, mid-size SaaS”) | Age and company size describe a market. They predict nothing about when someone looks. |
| Trigger stated as a desire (“wants to save time”) | Everyone wants to save time, permanently. Not an event, so there’s no moment to write to. |
| Skipping slot 6 because “we don’t want to exclude anyone” | Refusing to exclude is how copy gets general. The exclusion is what lets the rest be sharp. |
| Filling slots from a positioning doc | The doc is what you’re testing, not the source. Fill from customer conversations, support tickets, sales call notes, churn interviews. |
| Treating the card as documentation | It’s a gate. If it can’t be filled, the writing isn’t ready to start. |