Name the interactions being counted
Specify whether the numerator includes likes, comments, shares, saves or another available action. Do not add a metric the reporting source does not provide. YouTube’s Analytics reference distinguishes named measures; a general engagement label is not permission to invent a combined total.
Keep event count separate from distinct people. One person can perform several actions. A rate based on total interactions can exceed what a buyer expects if it is incorrectly described as the percentage of people who engaged.
Choose a denominator that answers the question
Follower-based rate compares interactions with the account’s following. View-based rate compares them with recorded views. Reach-based rate needs a suitable distinct-audience measure. They answer different questions and cannot be ranked as one interchangeable percentage.
For livestreams, do not casually use average concurrent viewers as if it were unique reach. Twitch documents current live viewer counts, but a snapshot does not establish all the people exposed to a sponsored segment.
Recalculate an invented post
Assume a fictional post records 300 counted interactions, 10,000 views and an account with 20,000 followers. Its view-based rate under that convention is 3%; its follower-based rate is 1.5%. Neither is automatically the correct universal engagement rate.
If reach is unavailable, leave the reach-based value unavailable. Do not estimate it from the other two merely to complete a spreadsheet. State the observation window and whether the counts include later activity after the original report.
Talk to us
Pick a 30-minute slot, or DM us on Telegram.
Need help applying this to your campaign? Tell us what you’re planning.
Pricing is scoped to your brief. We reply within one working day.
Compare distributions, not one selected winner
Review a relevant set of posts or streams with consistent definitions. Separate sponsored from other content when the task calls for it. A viral outlier or a post in another category may not describe the proposed commission.
Unusual ratios deserve questions, not instant fraud allegations. Check content context, source and reporting coverage. An apparently high rate does not prove audience eligibility, purchase intent or FTD performance.
| Component | Write explicitly |
|---|---|
| Numerator | The included interaction types and available source |
| Denominator | Followers, views, reach or another named measure |
| Window | Post age, reporting cutoff and later activity treatment |
| Aggregation | Relevant content set and handling of outliers |
| Limit | No inference of unique people, eligibility or acquisition without separate evidence |
Use the rate as one selection input
Combine the evidence with product fit, observed presentation skill and the actual campaign task. Keep unavailable data and confidence visible. Do not convert a good engagement percentage into a guaranteed acquisition forecast.
3betmedia can discuss an audience and creator review for a defined brief. Bring the source reports and the formula used in the deck. Recalculating the same measure is more valuable than comparing impressive percentages that mean different things.
Review average concurrency separately before using livestream audience figures as an engagement denominator.
Frequently asked questions
Which denominator should every brand use?
There is no single answer for every task. Name the question, source and available measure, then compare like with like.
Does a high engagement rate prove conversions?
No. It describes the chosen interaction measure under stated conditions, not customer acquisition or causal influence.
Results from campaigns our team has run
Web3 game · gaming creators, streams and events
Illuvium
6,550 new players and 24,000+ followers
Read the Illuvium caseCrypto exchange · KOLs on X, YouTube and TikTok
Binance
4,245 new users and more than $510M in trading volume, double the target
Read the Binance caseCrypto banking · influencers and growth
Swissmoney
32,000+ users onboarded and CPA cut 4x
Read the Swissmoney case
3betmedia launched in 2026; most of these campaigns ran before the name existed. See all 19 cases.
Sources and further reading
Primary references support the limited platform, product or rights facts above. Operational examples are hypothetical; this page does not provide legal approval or a performance guarantee.