Sample AI-visibility report · free
The report you hand the person who asked.
A full sample of the client-ready report, each section annotated with the question it answers.
A copy is on its way to . PDF · 702 KB.
PDF · 6 pages · 9 sections. Most of it is on this page; the whole sample comes as a PDF.
What’s inside
5 of 9 on this page. The rest, and the whole sample as one file, by email.
- 01Where you standFour numbers, each with its error bar.
- 02What they sayHow engines frame the brand when they name it.
- 03Against whomThe competitors engines name in the category.
- 04Engine by engineWhere the brand is strong, and where it is absent.
- 05The evidence trailThe sources engines take their facts from.
- 06AccuracyWhat engines get wrong about the brand, and the fix for each.In the PDF
- 07Proof of workWhat shipped, and what it measurably moved.In the PDF
- 08NextThe three moves for the next period.In the PDF
- 09MethodWhy the numbers hold up, in words a client can read.In the PDF
Northwind CRM — August 2026
7 days to Aug 4, 2026 · against the 7 before · 48 buyer questions × 4 engines, asked twice in the week · 384 answers in the window
The four numbers, with their error bars
✓ marks movement whose confidence intervals no longer overlap the window before. The +4 on Recommended stays marked ~ — it might be real, but the data can't prove it yet, and the report says it cannot rather than rounding up.
How engines frame Northwind when they name it
Who the engines name in your category
The first reading had Northwind fourth. The climb to #2 is the headline a client remembers — and it's defensible because the same 48 questions were asked on every slot.
Where you're strong, and where you're absent
| Engine | Mention rate | 95% CI | Trend over 21 days |
|---|---|---|---|
| Perplexity | 52% | 45–59% | rising — strongest surface |
| ChatGPT | 44% | 37–51% | rising since the pricing rewrite |
| Gemini | 33% | 27–39% | flat — the gap to work |
| Claude | 28% | 22–34% | flat, positive when present |
Where the engines get their facts
| Cited domain | Share of answers citing it | Names Northwind? |
|---|---|---|
| g2.com | 31% | Yes |
| reddit.com | 24% | No — names Bellrose & Keystone |
| capterra.com | 17% | Yes |
| wikipedia.org | 12% | Category page only |
The reddit row is where the next retainer month goes: the engines already trust it, and it currently tells them about two competitors.
What this sample cannot show you
Your numbers. Everything above is invented to show the shape — the real report is generated from readings of your own questions, and every figure arrives with the confidence interval those answers earned. The free checker won't produce this document, but it will answer the first question in it — whether you're in the answer at all — in about a minute: run it on your site.
Inside the product, this report opens from the dashboard in one click, shares as a no-login link for the person who asked, and can arrive by email on a schedule.
In the PDF
The sections that turn numbers into work are in the PDF.
Accuracy, proof of work, the next three moves and the method: the part of the report a client or a boss acts on. Every number in the sample is invented, and labelled that way.
- AccuracyWhat engines get wrong about the brand, and the fix for each.
- Proof of workWhat shipped, and what it measurably moved.
- NextThe three moves for the next period.
- MethodWhy the numbers hold up, in words a client can read.
A copy is on its way to . PDF · 702 KB.
- Use it as your template. Each section with the question it answers, ready to copy into your own report.
- Know when it changes. We email you when the template changes: a section added, retired or rethought, with the reasoning.
- Show it before you measure. Forward it to whoever asked, and agree what the report will cover first.
One email with your copy, then a note only when the sample changes. One-click unsubscribe on every email. How we use your address.