SocialHolmes

Finding out who is actually there

Follower and following lists and complete comment threads — the raw material for working out who an audience is rather than how big it is.

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.

Audience size is the least interesting thing about an audience

Follower count is one number and it answers almost nothing. The questions that change a decision — is this the audience we want, do these two accounts share one, is this community actually active — are all about composition, and composition means pulling the list.

Two sources are worth combining. Follower and following lists give you the nominal audience; comment threads give you the active one, and the gap between them is often the whole story. An account with a large following and a comment section of forty regulars is a different proposition from one with the same following and a thousand.

Overlap is the other thing worth computing and the thing nobody does, because it needs two lists rather than one. Two creators in the same niche can share most of an audience, which makes booking both of them considerably less useful than the follower counts suggest — and you can only find that out by pulling both and intersecting.

Can I compute overlap between two accounts?
Yes — pull both follower lists and intersect them. It is the single most useful thing in this category and almost nobody does it, because it needs raw lists rather than a dashboard summary.
Is a sample enough?
For ratios and overlap estimates, usually. A few thousand followers is enough to estimate composition with a sensible error bar, and it costs a fraction of the full list.
What about the privacy of the people in the list?
They are public accounts, but they are people. You are the controller of anything you store: set a retention period, have a lawful basis, and do not use this to build profiles of individuals. Our terms are explicit about it.