MentionBird can propose prompts for you. You'll meet suggestions in two places: during onboarding, when you're staring at an empty prompt list, and any time afterwards from your prompts page, when you want candidates you hadn't considered.
Where suggestions come from
Suggestions aren't a generic keyword list. They're built from what MentionBird already knows about your brand: your name and aliases, the competitors you've configured, the prompts you're already tracking, and the sources being cited in your category.
That's why they get more useful as your account matures — a brand with weeks of citation data behind it gets suggestions grounded in what assistants are actually being asked to compare in its market.
Built to produce data you can act on
Suggested prompts follow the same rules a good hand-written prompt does:
Brand-neutral. They ask the question a buyer would ask, without naming you — the whole point is to find out whether assistants bring you up unprompted.
Phrased as recommendations and comparisons. Those are the shapes that make an assistant go and search the web, and answers built on a live search carry citations — the data that tells you why an answer came out the way it did.
No year-stamps. A tracked prompt runs for months, and a year in the wording goes stale the moment the calendar turns.
No repeats. Anything you've already tracked — including prompts you've since archived — won't be suggested again.
When to use them
Two situations, one at each end of an account's life.
At the start, when you don't yet know what to track. Suggestions give you a defensible first set faster than a blank page does.
Later, when your prompts have plateaued. A tracked set tends to converge on the questions you already knew about. For Nike, suggestions might surface trail running or injury-recovery questions when the tracked set only covered marathon racing — adjacent territory a competitor may already own.
Judging what to accept
Treat suggestions as candidates, not a queue to accept wholesale. The test is the same as for any prompt: would a good answer move a real buyer closer to choosing? What makes a good prompt to track goes deeper on that judgement.
And remember that each accepted prompt consumes capacity on every platform you track it on — accepting ten suggestions across three platforms is thirty prompt runs. Accept the ones you'd be glad to still be tracking in three months.
