FAQ

Why does the first report take 15–20 minutes?

Your first report requires generating a custom semantic core of 100 prompts, sending them to all 5 AI models, waiting for complete responses, and running our full analysis pipeline — parsing, sentiment extraction, and scoring. After the initial build, weekly updates happen automatically in the background via batch processing, so you'll always see fresh data when you log in.

What does "blind test" mean?

We never ask AI models about your brand directly. Instead, we ask natural buyer-intent questions — the same ones your customers would ask — and check whether AI recommends you on its own. This prevents biased results and ensures the data reflects what real users experience.

Why does one model show 0% visibility?

This means that particular AI model never mentions your brand in response to any of your 100 buyer-intent prompts. It's not uncommon — each model has different training data and different behaviors. A 0% on one model is actually a useful finding: it tells you exactly where you have a gap. As that model updates its knowledge, your visibility may improve, especially if you optimize your web presence for AI discoverability.

Can I change which brand I'm tracking?

Yes, once per month. Go to Settings and select "RESET." This generates a completely new semantic core and runs a fresh scan, so it takes 15–20 minutes like the initial setup.

What counts as a "custom prompt"?

Custom prompts are additional queries you can write beyond the 100 automatically generated ones. They let you test specific scenarios or questions you care about. Base plans include 50 custom prompts per month, Pro includes 200, and Enterprise is unlimited.

Is my data private?

Yes. Your reports, scores, and competitor data are visible only to you and your team members. We don't share individual customer data with anyone. The prompts we send to AI models are generic buyer-intent questions that don't reveal who is doing the testing.

What AI models do you test?

Rankry currently tests ChatGPT (OpenAI), Claude (Anthropic), Gemini (Google), Perplexity, and Grok (xAI). All models are tested with live web search enabled to match real-world usage conditions. We plan to add more models as the market evolves.

Can I export my data?

All plans include data export capabilities. You can download your reports as CSV or PDF from the dashboard. Enterprise plans also include API access for integrating Rankry data into your own tools and workflows.

Updated on: 02/03/2026