Vetting an influencer on their audience, not their media kit
Follower and following lists, and the comment threads under recent posts. The things a bought audience cannot fake, in a shape you can run checks against.
What serves this
Read from the platform registry, so nothing here is a job the API cannot currently do. Each endpoint links to its own reference.
What a bought audience cannot hide
A media kit is a claim, and follower count is the easiest number in marketing to buy. What is much harder to fake convincingly is the shape of an audience and the texture of a comment thread — which is why vetting works from follower lists and comments rather than from headline figures.
The signals worth computing are all ratios and distributions. Follower-to-following on the followers themselves; how many of them have no posts and a default avatar; how concentrated engagement is among a handful of accounts that comment on everything; whether the comments say anything or are three words and an emoji. None of these is conclusive alone and all of them together are hard to manufacture.
This is data about real people, so it comes with obligations. You are the controller of whatever you store, and follower lists in particular are personal data in most jurisdictions — worth having a retention period and a lawful basis before you build the pipeline rather than after.