Open any prompt and you get a leaderboard: every brand named in the answers to that question, ranked by how often. It's the most direct answer to "who is winning this question, and by how much?"
What each column means
Times named — how many separate checks mentioned that brand.
Mention rate — the same figure as a percentage of all checks on this prompt.
Avg position — where the brand lands when an answer ranks brands in order. Blank if it has never been ranked. Only ranked answers feed this, so an in-text mention doesn't drag the number down. See the metrics article.
Which platforms named it — often the most actionable column, since a competitor dominating one assistant and absent from another usually points to a specific source that assistant trusts.
Your brand isn't pinned to the top
Your row sits at its true position in the ranking. If Nike is fourth on a question, it appears fourth.
This is deliberate, and worth knowing before you go looking for yourself at the top. If you're outside the top ten, your position is shown separately below the table — so the honest ranking stays intact without you having to hunt for your own row.
Variant names are collapsed
Mentions of Nike, Nike Air Max and NikeCourt count toward one row, not three. The same applies to each competitor you've configured.
Within a single answer, several variants count as one mention. An answer discussing three Nike models isn't three mentions — otherwise brands with many product names would look artificially dominant.
Brands you didn't name
Any brand appearing in the answers gets a row, whether or not you've configured it as a competitor. That's often where the surprises are: the leaderboard shows the shortlist assistants actually offer, which may not match the one in your internal deck.
Finding competitors to add
MentionBird can suggest competitors based on brands that keep appearing alongside yours.
The ranking behind those suggestions is worth understanding, because it's not simply "most mentioned". Raw mention counts are dominated by ecosystem names — marketplaces, platforms, retailers — which get mentioned in passing constantly without being anything a buyer would choose instead of you.
So suggestions are ranked by structured mentions: appearances in ranked or bulleted lists, the "top alternatives" shape. That's where real competitors show up, while platforms tend to appear in flowing prose. The effect is that a genuine rival mentioned in ten shortlists outranks a marketplace mentioned fifty times in passing.
Suggestions are also filtered to drop things that aren't alternatives to you — publications, review sites, communities, general-purpose tools used alongside your product. The filter errs toward keeping borderline cases, on the basis that you're reviewing each one anyway.
Nothing is added until you confirm it. Suggestions are candidates; adding a competitor is always your decision.
Reading it well
The single most useful comparison is your mention rate against the leader's. Being named in 20% of answers where the leader is named in 80% is a specific, measurable gap — and the sources cited on that prompt are where the explanation usually lives.
