A finding count is a numerator. On its own it carries no information: forty flagged posts is a rounding error on one estate and a crisis on another.
The obvious fix is to state the denominator — how many posts were enumerated, over what period, on which accounts. Every serious methodology says to do this. Far fewer acknowledge how hard the denominator is to establish, because on several major platforms the number is not published at all.
What each platform will actually tell you
The position differs sharply by platform, and it is worth knowing before you commission anything:
- Instagram — states an exact post count in the profile header. Read it and move on.
- YouTube — gives a video count on the channel. Shorts are frequently counted separately from long-form, so one number may not be the whole channel.
- Facebook — an exact figure is available through a Business Suite export, which requires admin access to the page. Without it, you are counting by hand.
- TikTok — publishes no total. The grid has to be fully loaded and the thumbnails counted.
- Pinterest — publishes no total for created pins.
- LinkedIn — publishes no post count either.
Two platforms hand you the number. One hands it over only to an administrator. Three do not have it on offer in any form.
So the denominator is constructed, not read
On more than half your surfaces, the figure in the “posts reviewed” line of a report was produced by a method: an export, an enumeration, or an estimate. Those are not equivalent, and which one was used changes what the resulting percentage can be asked to support.
The method therefore belongs in the report next to the number. An estimate that is labelled an estimate is a useful, honest input. The same estimate presented as a count is a figure that will not survive being questioned — and it will be questioned at precisely the moment it matters, by somebody whose job is to find the soft part of your evidence.
The failure this produces
A rate calculated from a constructed denominator, presented as though it were counted, is the single most common way an otherwise sound audit becomes indefensible. The findings may be perfect. The percentage sitting above them in the summary is the part that gets challenged.
Why this decides whether a finding count means anything
Two reports both say “we found 40 issues”. One enumerated 600 posts; the other enumerated 40,000. They describe entirely different situations, and nothing on the face of either document distinguishes them unless the denominator and its provenance are stated.
This is also what makes findings comparable over time. A remediation programme is judged on whether the rate moved, and a rate whose denominator was built by a different method each quarter cannot show that. Scope drift masquerades as progress.
What to ask
Of any audit, internal or external, before you read a single finding:
- What was the denominator, per platform? Not a total across the estate — the platforms behave differently and the number is trustworthy on some and constructed on others.
- How was each one established? Read from the platform, exported with admin access, enumerated, or estimated.
- What was excluded, and why? Private accounts, deleted posts, surfaces that were out of scope.
- Will the same method be used next time? If not, the two reports do not compare.
A provider who answers these quickly has done the work. A provider who treats the denominator as a formality has told you how much weight the percentage can bear.
What else belongs alongside the denominator in a document that has to survive scrutiny is set out on the methodology page.
Not sure what is sitting on your own channels?
Send us one Instagram, TikTok or YouTube profile and we will check its recent posts for unlicensed music, then send you a private risk report. No account access, no card details.
