Your product ships every sprint. When did your website last ship?
AI agents now research, compare, and sign up on behalf of your customers. Find out if they can find your site, read it, use it, and call it. About 60 seconds.
Before you scanGuess your score.
Confident. We will put your guess next to the real number.
Visibility is one layer. We test all five.
Most AI checkers stop at "do chatbots mention you." We send a real agent to your site and see what it can actually do.
- 01
Findable
Do AI answers cite you?
Sampled prompts for your category, structured data, sitemap.
Who fixes it- COLAB, fixes this
- Pantheon, not needed
- 02
Readable
Can agents get your content?
Crawler rules, CDN bot blocking, JS-only pages, llms.txt.
Who fixes it- COLAB, fixes this
- Pantheon, fixes this
- 03
Usable
Can an agent finish a task?
A real browser agent runs your top buyer tasks and logs each step.
Who fixes it- COLAB, fixes this
- Pantheon, not needed
- 04
Callable
Do you expose tools to agents?
WebMCP, MCP endpoints, public APIs, schema.org Actions.
Who fixes it- COLAB, fixes this
- Pantheon, fixes this
- 05
Governed
Did anyone decide the rules?
AI policy, intentional vs accidental blocking, permissions.
Who fixes it- COLAB, fixes this
- Pantheon, fixes this
The usability test you never ran
You test with users. Nobody tested with agents.
We give an agent three jobs your buyers would give it, then log every step. Where it gets stuck is usually where a screen reader user gets stuck too, so the fixes pay off twice.
task: "find pricing and start a free trial"
- 01open homepage: ok
- 02locate "Pricing" in nav: ok
- 03read plan table: ok
- 04click "Start free trial": ok
- 05fill signup form: 2 fields unlabeled
- 06submit: blocked by CAPTCHA
result: failed at step 6 of 6
Your roadmap has a backlog. Now your website does too.
Every report ends with ranked fixes, sized like tickets. Hand them to whoever owns the site, or bring them to both of us. COLAB fixes the site. Pantheon runs it.
Check your site