Why does every tool give me a different engagement rate?

Because they are using different formulas — by followers, by reach or by impressions — and often counting different actions as an engagement. Check which one a tool uses before comparing its output to anything.

Do saves and shares count?

It depends who is counting. Most platform-native analytics include them; many third-party tools computing from outside cannot see them at all, which is one reason external estimates run low.

What is a good engagement rate?

Higher than your own recent median. Published benchmarks disagree with each other and are computed on samples that are unlikely to resemble your audience.

Engagement rate, and why yours looks different everywhere

Three formulas share one name. Any two tools will disagree about your engagement rate, and both can be right.

Last reviewed 31 August 2026

The three formulas

All three are called "engagement rate". They measure genuinely different things, which is why your number changes depending on who calculated it.

By followers: engagements ÷ followers × 100. The one most brands quote, because followers is the only input they can see from outside. It punishes you as you grow, since follower count rises faster than the share of them any single post reaches.
By reach: engagements ÷ accounts reached × 100. The most honest measure of whether the content worked, because it only counts people who actually saw it. Needs your own analytics — nobody can compute it from outside.
By impressions: engagements ÷ impressions × 100. Always lower than the by-reach figure, because impressions count the same person more than once. It is not necessarily the lowest of the three — by-followers sits below it whenever your follower count exceeds the post's impressions, which for most accounts is the normal case.

Which one to quote, and to whom

Quote by-reach in a pitch and say so explicitly, because it is the number that describes your content rather than your follower history. Expect the brand to also compute by-followers themselves, and expect it to be lower. Neither of you is wrong.

What matters far more than which formula: use the same one every time you compare anything. A rise from 2% to 3% means nothing if the first was by reach and the second by followers.

Why cross-platform comparison is meaningless

Each platform counts a different set of actions as engagement, and each shows content to a different share of your followers by default. A TikTok rate and an Instagram rate are not on the same scale, and the ratio between them says more about the two algorithms than about your content.

Compare a platform against itself over time, and compare your posts against your own median. That is the only comparison that carries information.

What counts as good

Published benchmarks vary enormously between sources, and most are computed from whatever sample of accounts the publisher happened to have. Treating any of them as a target is how creators end up chasing a number that was never measured on an audience like theirs.

The benchmark worth having is your own median over the last thirty posts. A post above it worked; one below it did not. That comparison is measured on exactly the right sample — your audience — which no published table can claim.

Questions

Why does every tool give me a different engagement rate?: Because they are using different formulas — by followers, by reach or by impressions — and often counting different actions as an engagement. Check which one a tool uses before comparing its output to anything.
Do saves and shares count?: It depends who is counting. Most platform-native analytics include them; many third-party tools computing from outside cannot see them at all, which is one reason external estimates run low.
What is a good engagement rate?: Higher than your own recent median. Published benchmarks disagree with each other and are computed on samples that are unlikely to resemble your audience.