1. Set up your brand
Your domain in, a brand profile out, confirmed by you.
The setup reads your site and drafts a brand profile: your name and the other ways people write it, your category, your audience and your pricing facts. You correct it and add a short truth file of 5 to 15 facts, the things an AI answer should get right about you. Competitors are suggested from search data and corrected later by the brands the first test actually names.
2. Pick the prompts your buyers ask
20 tracked prompts across the funnel, plus two controls.
Prompts are generated from templates for each stage of the buying journey and labeled as branded, pricing, comparison or discovery. You add the 3 to 5 questions you most want to win. Each prompt shows whether engines usually search the web to answer it, because a question answered from memory will not move when you change your site. Two control prompts that no fix will target keep the trend line honest.
3. Test on four engines, 10 runs each
Headline numbers come only from consistency tests.
A consistency test asks one prompt 10 times on each engine (3 times on Google AI Overviews) and labels the result STRONG, MODERATE, WEAK or INVISIBLE. Your mention rate per engine, your share of voice against each competitor and the cited domains ranking come from these tests. A weekly spot check of every prompt runs in between and is shown as one sample, never as a score.
Every run stores the full answer, the cited sources, whether the engine searched, and the engine version, so any number on the dashboard can be traced back to the answers behind it.
4. Grade and rewrite your own pages
The checks that decide whether an engine can quote you.
- First 50 Words: does the opening say who you are, what you do, where you work and what it costs?
- JSON-LD: is structured data present and valid?
- Robots rules: are the AI crawlers allowed in?
- llms.txt: does the site publish one?
- Server rendering: can a crawler that does not run JavaScript read the copy?
- Category words: do the title and meta description name what you sell?
Every page that fails the First 50 Words check gets a rewritten opening built from your truth file, shown before and after, ready to copy.
5. Work the playbook
Findings turned into a list of jobs, in order.
The playbook turns every finding into an item. A prompt where engines search but never name you becomes a page to create. A page that fails the grader becomes a page to fix. A domain that engines cite, that names your competitors but not you, becomes a source gap for your team to work on.
6. Log the fix and prove it
A dated entry for every change, then a new 10 run test.
When you ship a fix, you log it with the date and the prompts it targets. The next consistency test shows the change next to the log entry, so you and your client can see which fix moved which answer, and which did nothing. Reports share at a private link, with a print version for PDF.
Questions people ask
Short answers, in the words buyers use.
How long does setup take?
A few minutes to confirm the brand profile, the truth file, the competitors and the prompts. The first consistency tests then run in the background.
Do I need to change my website to use mentionedin.ai?
No. Measuring needs only your domain. The page grader and the rewrites tell you what to change, and you decide what ships.
What does mentionedin.ai do with pages where my competitors are listed and I am not?
It lists them as source gaps in the playbook, ranked by how often engines cite them, so your team knows which third party pages to work on. Nothing is sent to those sites on your behalf.