The formula has two choices in it
Every engagement rate has the same shape: the interactions you count, divided by a measure of audience, usually multiplied by 100. The arithmetic is trivial. The decisions are not: which interactions go in the numerator (likes only, or likes, comments, shares and saves), and which audience measure goes underneath.
Two people can report a '4% engagement rate' for the same post and both be right, because one divided by followers and the other by views. Until both choices are written next to the number, the number cannot be compared with anything.
Views, reach, or followers: what each one answers
Each denominator answers a different question. Pick the question first.
| Denominator | Question it answers | Watch out for |
|---|---|---|
| Views | Of the times this post was shown or played, how often did someone respond? | A view is counted by each platform's own rules, and one person can view a post more than once |
| Reach (accounts) | Of the distinct accounts that saw it, what share responded? | Only meaningful where the platform reports deduplicated accounts; reach on one platform cannot be added to reach on another |
| Followers | How much of my existing following responded? | Followers are not viewers: many followers never see a given post, and posts that travel beyond your followers can exceed 100% |
Why follower counts make a weak denominator
YouTube's own help page makes the point for its platform: subscriber count is not the most accurate way to estimate the size of your active audience, because viewers are usually subscribed to many channels and may not return for every new upload. YouTube instead offers a monthly audience, the unique viewers who watched in the last 28 days.
The same logic applies elsewhere. A follower-based rate mixes how good the post was with how many of your followers the platform chose to show it to. It is still useful for one question (is my following engaged?), but it is a poor way to compare individual posts, especially short videos that are mostly seen by people who do not follow you.
Synthetic example: the same three posts, two rankings
These numbers are invented to show the arithmetic; they are not customer data or a benchmark. One creator publishes three posts in a week, on three different accounts with different follower counts. Interactions are likes, comments or replies, shares or reposts, and saves where shown.
| Post (synthetic) | Interactions counted | Views | Followers | Rate per view | Rate per follower |
|---|---|---|---|---|---|
| Short video | 240 likes + 30 comments + 30 shares = 300 | 12,000 | 8,000 | 2.5% | 3.75% |
| Text post | 90 likes + 45 replies + 15 reposts = 150 | 1,500 | 2,000 | 10% | 7.5% |
| Carousel | 120 likes + 10 comments + 20 saves = 150 | 4,200 | 5,000 | 3.6% | 3% |
What the example shows
Divided by views, the carousel (3.6%) beats the short video (2.5%). Divided by followers, the short video (3.75%) beats the carousel (3%). Nothing about the posts changed; only the denominator did. If the carousel's platform also reports 3,000 accounts reached, its rate per account reached is 5%, a third figure for the same post.
The numerators differ too. The carousel's count includes 20 saves; the text post's platform shows no saves, so it could not earn any. Comparing them is defensible only if you say so, or if you restrict the numerator to the interactions every platform reports, such as likes and replies.
When a metric is missing
Platforms do not all expose the same fields, and some only show certain metrics to the account owner or above a threshold. When a denominator or interaction type is unavailable, leave the cell empty and mark it unavailable. Do not substitute followers for missing views, and do not treat a missing value as zero; either one quietly changes the formula for that row.
- Write the formula at the top of the sheet: which interactions, which denominator, which date range.
- Use each platform's own definition of a view or reach, and note where the definition lives.
- Mark unavailable metrics as unavailable, not zero.
- Compare rates only between rows that use the same numerator and denominator.
- If you need one cross-platform figure, report summed interactions and summed views as totals, labeled as platform-reported views, not as unique people.
Be careful with published 'average engagement rate' figures
A benchmark is only useful if it states its formula, the platforms, the account sizes, and the period it covers. A by-follower average compared with your by-view rate tells you nothing. The most reliable baseline is usually your own: the same formula applied to your own recent posts on the same platform.
Where Ferryman fits
Ferryman's analytics view brings supported views, likes, replies, and reposts from posts published through Ferryman into one place, with per-platform comparisons and top-post rankings, refreshed on a recurring six-hour cadence. That gives you the views and interactions for a by-view rate without building the spreadsheet by hand. The analytics dashboard requires an active paid subscription.
Ferryman does not report deduplicated people across platforms, and it does not decide which formula you should use. Choosing the denominator is still your call, and the rule above still applies: one formula per comparison.
FROM GUIDE TO WORKFLOW
Explore this workflow in Ferryman
See supported engagement across platforms
Review supported views, likes, replies, and reposts from posts published through Ferryman, compare platforms, and open the posts behind the totals.
Sources & further reading
Written by Ferryman. Sources checked Nov 2, 2026. Platform rules can change; consult the linked documentation for current requirements.