1. The 2026 formula
This is the formula every serious Instagram analytics tool has converged on in 2026. It's the one Meta itself uses internally when surfacing insights, and the one creator marketplaces use when ranking creators.
The older formula — just likes and comments divided by followers — systematically under-reports your true engagement because it ignores saves (the highest-weighted signal in Meta's algorithm) and over-inflates the denominator with inactive followers who never see your posts.
2. Why saves suddenly count
Saves didn't matter in 2022 because the API didn't expose them reliably. That changed in late 2024 when Meta opened up saves data to third-party analytics platforms. Immediately it became obvious that saves were the most valuable engagement signal — users who save a post are 5–10× more likely to come back to the account and convert on a sponsored product.
If you're still calculating ER without saves, you're using a 2022-era definition. Update the formula and expect your reported ER to jump 15–25%.
3. Reach vs followers — which denominator?
Reach measures who actually saw your post. Followers measures who could theoretically have seen it. In 2026, the gap between those two numbers is enormous — typical Instagram accounts reach only 15–30% of their followers on any given post.
If you don't have access to reach (public accounts can't see competitors' reach), fall back to followers and expect the number to read 2–4× lower than the true reach-based value. Agencies treat the reach-based formula as the only legitimate one.
4. ER by post format
Carousel ER is 1.5–2× higher than static post ER across every niche we measured. The mechanism: each swipe counts as engagement, and Instagram's algorithm rewards multi-swipe behavior by pushing the post to more people. Posting a 10-slide carousel is mathematically equivalent to posting 10 static posts in terms of signal strength — without spamming the feed.
Reels ER is lower per-view than feed post ER but reach is dramatically higher. On a per-post basis they net out roughly equivalent to carousels. Static posts underperform both by a wide margin and should represent no more than 20% of your posting cadence.