MentionBird tells you whether AI assistants mention your brand when people ask questions in your category — and where those assistants got their information.
Five ideas explain every screen in the product. Once these make sense, the rest of the dashboard reads easily.
We'll use the same example throughout this help center: imagine you work at Nike, and you want to know whether AI assistants recommend Nike when people ask about running shoes.
Prompt
A prompt is a question you want to monitor — something a potential customer would actually type into an AI assistant. In our example:
What are the best running shoes for marathon training?
Notice what's missing: the word Nike. That's deliberate, and it's the most important habit to build. Good prompts are brand-neutral.
If you tracked "are Nike shoes good for marathons" instead, you'd only learn that an assistant can describe a brand you handed it. You'd learn nothing about whether Nike comes up on its own — which is the thing that actually wins customers.
AI platform
Each prompt is tracked separately on each AI platform, because different assistants answer the same question differently. Ask that marathon question on three assistants and you might get three different shortlists — one leading with Hoka and Brooks, another mentioning Nike third, a third not mentioning Nike at all.
That's why one prompt tracked on three platforms is three separate things to check. The platform picker shows which platforms are available on your plan.
Run
A run is one check: one prompt, asked on one platform, once. MentionBird performs these on a schedule and records the full answer every time.
Answers vary between runs even when nothing about your brand has changed. Ask about marathon shoes twice in a week and the order may shift. That's how these models behave, not a fault in the data — and it's exactly why tracking over time matters. One answer is an anecdote; the same prompt checked repeatedly becomes a trend you can act on.
Search
To answer a prompt, an assistant often searches the web first — and what it searches for is its own wording, not yours.
Our marathon question might send it looking for best marathon running shoes, carbon plate running shoe reviews, and Hoka vs Nike marathon. Those searches decide which pages the assistant ever sees, and so which pages can end up cited.
Not every answer involves a search — sometimes the assistant replies from what it already knows. Both are recorded, and the difference matters: a prompt answered from memory can't be moved by a page you publish today. See What the AIs searched for.
Citation
When an assistant references a web page in its answer, that's a citation.
Our marathon answer might cite a Runner's World roundup, a Reddit thread comparing carbon-plated shoes, and a YouTube review. Those pages are what shaped the recommendation.
Citations are the most actionable data in MentionBird, because they show which pages are really deciding what AI says about your category. Some will be yours. Some will be competitors'. Many — like that magazine roundup — belong to nobody you control, and knowing which of those carry weight tells you where to focus.
How they fit together
Prompts multiplied by platforms is what gets checked. Each check is a run. Runs produce brand mentions, the searches the assistant ran, and page citations.
Everything else in MentionBird — visibility scores, competitor comparisons, source leaderboards, ranking advice — is those runs aggregated in different ways. So when a number looks surprising, it's usually worth going back to the individual runs behind it.
Where to go next
Setting up your first brand — get tracking started
What makes a good prompt to track — the highest-leverage decision you'll make
Understanding your first results — what's worth looking at on day one
Glossary of AI visibility terms — every term in one place
