Hook: Hand a writer the list of buyer questions your own pages never answer. Scoring that list by hand costs EUR 540 to EUR 720 a round, and it is out of date the week you publish.
What You'll Be Able To Do
- Read a coverage board with your buyer types down the side, the topic clusters found on your own website across the top, and one number in every cell
- Open any buyer type and pull its Uncovered needs list: real questions, in a customer's own words, that your pages do not answer
- Read the mirror image, the questions your pages already answer, so you stop rewriting the pages that work
- Rank the same gaps across the whole audience, sorted by how many buyer types share each one
- Send a single gap onward with the buyer and the question already chosen
- Export a Persona Bank CSV, fifteen columns, one row per buyer type with its AI Coverage figure
Runtime: 5:13 | Difficulty: Intermediate
You own the content calendar. Every quarter you decide which eight things get written, and you decide it from keyword volume, a competitor's blog and the loudest voice in the meeting. The job goes wrong three months later, when the traffic report lands and nobody in the room can say which of the eight articles changed a single answer about your brand.
What it costs today
Five buyer types scored against roughly 15 content clusters, so 75 judgement calls, one round.
| Line item | Arithmetic | Cost |
|---|---|---|
| The scoring round, in-house | 6 to 8 h at EUR 90/h | EUR 540 to EUR 720 per round |
| The same round at an agency | 6 to 8 h at EUR 120/h | EUR 720 to EUR 960 per round |
| Bought in as a content-gap audit | 1 to 3 weeks, delivered once, never refreshed | EUR 2,000 to EUR 8,000 |
What the in-house round buys you is one person's opinion about what your pages cover, and that opinion goes stale the week you publish. One round costs what the Marketer plan costs in three to four months, and the plan recomputes the whole map every time you re-scan the website or change the audience definition.
What the video shows
- Cluster Coverage opens the board. Rows are your buyer types, columns are the topic clusters the crawler found on your website, and each cell is one number: how close that person's questions sit to the pages you published. Teal means covered, red means a hole.
- The cursor tracks slowly along the reddest row. The other rows fade back and a tooltip follows the pointer with the person, their role and the first three gaps behind that score.
- One click opens the person. What they are trying to get done sits on the left, the questions they put to a model sit on the right, phrased the way they would phrase them.
- Uncovered needs is the list you hand over. Every line is a question this buyer asks that your website does not answer, with exact duplicates stripped out, so counting the lines gives you your next brief list.
- Covered needs is the mirror image. Per topic, the coverage number and the questions your pages already answer, in cards that fold shut when you have read them.
- The Overview ranks the same gaps across everybody. Top Pain Points keeps the top three, each with the count of buyer types that share it, and each with a Fix it pill that opens the content generator with that person and that gap already chosen.
- Persona Bank, Export, CSV. Fifteen columns: role, seniority, buying journey stage, decision weight, AI Coverage and the queries they type.
- The closing puts the arithmetic on screen. Six to eight hours of judgement calls at ninety euros an hour, against a plan at a hundred and ninety-six euros a month.
What you need
- Marketer, EUR 196 a month. Below that tier the coverage screens render empty.
- One finished scan on the project. The model picker offers Gemini, ChatGPT and Grok, and any of the three works here.
- A curated Persona Bank of four to six buyer types. Four is enough to read a pattern, and above six the board stops fitting on one screen.
- Every persona edit made before you read the numbers, then Re-measure pressed and finished. Editing a persona resets its coverage to "Not measured yet", so a late edit blanks the row you were about to present.
Re-measure takes a few minutes on a six-persona project and shows a banner while it runs. Wait for it to clear before you screenshot anything.
Questions people ask
Is the coverage number a comparison against a competitor? No. It measures your own pages against your own buyer types. Nothing in that number reads a rival's website.
Do the gap sentences come out in the export? No, and this is the limit that catches people. The CSV, Markdown and PDF stop at AI Coverage. The CSV does carry a Pains column, which holds the persona's declared pains, a different thing from the computed gaps. Screenshot the block, or send a single gap onward with Fix it.
Do discovered personas get a coverage number? No. Coverage is computed for the curated Persona Bank only. Discovered personas stay on their own tab with no number against them.
What happens to a persona with no number? It reads "Not measured yet" in plain words until Re-measure finishes, so you never hand a board with a fabricated zero on it to a client.
How often can I refresh it? Re-measure is free and carries no limit, so press it whenever you change the audience definition. The website scan behind it runs once per model per calendar month. There is no schedule and no alert here; the recompute happens when you press the button.
Watch the five minutes, then hand the Uncovered needs list to a writer.