How we collect and measure creator data

Every number on infloo — in the rankings, the free tools, the fake follower check and the engagement rate benchmarks — comes from public social media data and a method we can describe. This page says where the data comes from, how fresh it is and how each metric is calculated, so you can judge how far to rely on it.

Where the data comes from

We work only with public data: what anyone can see on a public Instagram or TikTok profile — the follower and following counts, the posts, and the likes, comments and views they get (on TikTok, shares too).

The rankings come from infloo’s database of creators, grouped by country, city and niche. A ranking page lists the largest accounts of its place and niche by followers, each with its engagement rate.

The fake follower check collects its own sample the first time an account is checked — the accounts that liked its latest posts — and keeps the report until a data update collects it again. A private account hides its posts and likes, so the check covers public accounts only. The benchmarks are built from the rankings themselves, once a month.

How fresh the data is

Each kind of data is renewed on its own schedule:

Data How it is renewed
Rankings by country, city and nicheThe lists are downloaded again every week; a category ranking names the month of its list.
Engagement rate benchmarksA new snapshot once a month, in its first days; the page gives its date, the CSV its month.
Follower count historyUp to the last time we collected the account; for a long history the growth chart shows averaged points, not every day.
Fake follower checkCollected the first time an account is checked; a data update collects it again.

A profile’s follower count and engagement rate are as of the last time we collected that profile.

Engagement rate

Engagement rate (ER) shows how many people react to an account’s posts relative to the size of its audience. infloo shows three kinds of it, and they do not compare with one another:

  • Profile engagement rate — the average number of interactions on an account’s recent posts, relative to its followers. This is the ER in the rankings, the benchmarks, the engagement rate calculator and a report’s profile numbers.
  • Engagement rate in the Infloo Score — the average likes plus the average comments on the posts the fake follower check reads, divided by the account’s followers. It is the engagement part of the score, so it can differ from the profile engagement rate.
  • TikTok engagement rate — likes, comments and shares on an account’s last 12 videos, divided by their views. It is measured against views, not followers, so it does not compare with Instagram’s.

Large accounts get a lower rate, so compare an account with others of the same size in the same country and niche: the engagement rate benchmarks give the medians.

The engagement rate calculator shows the profile engagement rate of any public Instagram account.

Infloo Score

The Infloo Score rates the quality of an account’s audience from 0 to 100. It adds up three parts: 50% for authenticity — the share of real people and influencers among the accounts that react; 40% for engagement — the engagement rate of the posts the check reads, measured against 6%, the level at which this part is full; and 10% for how regularly the account posts. The total falls into one of six bands:

Band Score
Excellent90–100
Very good80–89
Good60–79
Average40–59
Poor20–39
Very poor1–19

The sample, each group of accounts and what the check cannot see are described in full on the fake follower checker page.

Fake followers and audience quality

Instagram does not show anyone a large account’s full follower list, so the check reads a sample of the accounts that liked the last 5 posts, and looks up the follower and following counts of up to 170 of these accounts. It sorts the accounts into four groups:

  • Real audience — what is left once the other three groups are taken out.
  • Influencers — accounts with 5,000 or more followers of their own.
  • Mass followers — accounts that follow 1,500 or more profiles, and at least 2 times as many as follow them. Their feed is so crowded that few posts reach them.
  • Suspicious — accounts that look like bots: no profile photo, no display name or a name with no vowels, a username with many digits. One sign alone is not enough; it takes several together.

This is the full report. For an account that has no report yet, this site runs the quick Lite check, which has no sub-sample: it counts the verified accounts among the likers as influencers and leaves mass followers out.

Bought followers who never like anything do not appear among the likers, so the sample cannot count them. They show up elsewhere: as an engagement rate far below that of accounts of the same size, and as a sudden step on the follower growth chart.

Engagement rate benchmarks

The Instagram benchmarks are medians of the profile engagement rate across the creators in our rankings: in each country, the 50 largest accounts overall and the 50 largest in each niche except the adult ones, each counted once. The TikTok rows, where a country has them, are medians of TikTok’s engagement rate by views across the country’s TikTok top. A row is published only when at least 10 of its accounts — and at least half of all its accounts — have an engagement rate. Accounts without one count only toward the sample size and the median followers. A new snapshot is taken every month and keeps its date.

How the sample, the coverage and each median are worked out: the benchmarks’ methodology.

What the numbers can and cannot tell

Every audience figure is an estimate from a sample, not a count of every follower: the sample is the accounts that liked the posts, so the fewer likes an account gets, the less its figures say.

Other services read other samples and weigh signals differently, so their figures for the same account will differ from ours. Compare accounts within one tool, not across tools.

If a number looks wrong, write to support@infloo.io with the account and the page, and we will look into it.

Who prepares the data

The infloo data team builds and runs infloo’s data collection, the rankings, the fake follower check and the benchmarks, and publishes the infloo blog. Articles are signed by the team, not by one person; the ones built on our data quote figures measured by the methods on this page. When a method changes, this page changes with it and its review date moves.

Free tools