Engagement rate by followers vs by reach
Two ways to measure engagement rate, why they diverge, and which one to report so your social numbers stay honest as reach shrinks.
Two denominators, two different stories
Engagement rate always divides engagements by a base, but the base you pick changes the meaning of the result. Dividing by followers answers how much of your total audience reacted, which is the headline number most brands and sponsors expect to see. Dividing by reach answers how the people actually served the post responded, which isolates content quality from distribution. A post can score 3% by followers and 12% by reach at the same time, and neither number is wrong.
Why the two rates keep drifting apart
On most networks organic reach is now a fraction of follower count, so far fewer people see a post than follow you. As that gap widens, the followers rate falls even when your content lands well, because the denominator stays large while views shrink. The reach rate holds steadier because its denominator moves with actual views. Watching both side by side tells you whether a soft week came from weaker content or weaker distribution.
Which rate to report and when
Use the followers rate for public benchmarks, media kits and sponsor pitches, since it is the figure outsiders can verify from your visible follower count. Use the reach rate for internal reviews and A/B tests, where you care about which creative earned the strongest reaction from the people who saw it. Whichever you choose, label it clearly, because comparing a followers rate against someone else's reach rate is meaningless.
Keep the inputs consistent
Comparisons only hold if you count engagements the same way every time. Decide up front whether saves, shares and profile taps are in or out, then apply that rule to every post. Pull reach and follower numbers from the same window as the engagements, and avoid mixing lifetime engagements with a snapshot follower count. Consistent inputs are what let a two-decimal percentage actually mean something across months.