What this guide is really about
You posted on Threads, the likes trickled in, and now you want a way to compare its response with other posts. Fair enough. Raw counts lack context. A rate adds a denominator, though it still won't tell you everything about the post.
The math takes ten seconds. What trips people up is what goes under the division line, where the counts actually live in the app, and what quietly skews the result once you have it.
By the end you'll have both formulas, where to find the post counts and which other input you'll need, and a way to judge the number against your own history instead of somebody else's benchmark.
Engagement rate on Threads = (likes + replies + reposts) / views x 100. Divide by followers instead when you're comparing accounts of different sizes. Open your profile and post to find the interaction counts below it; use View activity for details about who interacted. Record views and, for the follower-based formula, your follower count. Compare rates calculated with the same denominator and at the same post age.
Both formulas and when each one applies
Where to find every count inside the app
Four traps that quietly skew the math
A weekly logging routine you'll actually keep
Threads engagement rate = (likes + replies + reposts) / views x 100 for post response on the same account.
By followers: (likes + replies + reposts) / followers x 100. Useful for comparing interactions relative to follower counts across accounts of different sizes.
Reply counts on a thread you start include all replies in the whole thread, not just direct replies to your post, which can inflate the top line.
Views count across Threads and Instagram when a post is shared, so a view-based rate is not a Threads-only number.
The only fair benchmark is your own trailing baseline at the same post age and the same denominator.
Pick the Denominator Before You Touch the Numbers
Before any division happens, decide what goes under the line. Threads gives you two sane choices, views or followers, and the top line never changes: likes plus replies plus reposts.
Divide by views when you want to know how the people who saw the post responded. That's the right lens for comparing one post against the next on the same account. Divide by followers when size gets in the way, because a bigger account banks more raw views, and a follower-based rate expresses interactions relative to each account's follower count, but it does not account for differences in reach. One more denominator shows up in the wild: impressions, the total number of times a post was displayed. Threads surfaces views, not impressions, so treat any impressions-based rate from another platform's study as a different animal and never paste it into a Threads comparison.
Then one rule with teeth: never present follower-based and view-based rates in a single comparison. Name the denominator every time you quote a rate, in your notes and out loud. Do that and the number becomes something you can act on, a baseline for ways to increase your engagement rather than a trivia answer.
Find Each Number Inside the Threads App
Open your Profile, then the post. The counts below it show likes, replies and reposts. Select View activity to see who liked, reposted or quoted it. Record the view count separately for the view-based formula. In the mobile app the same numbers sit on the post itself: tap the post to open it full-screen and the like, reply and repost counts sit under the text, with the view count on the post and a View activity option for the who-did-what list.
Don't treat View activity as a single screen containing every formula input: Meta describes it as the place to see who liked, reposted or quoted a post. One wrinkle to know early: views count across Threads and Instagram when a post is shared, so a view-based rate may include exposure on Instagram as well as Threads.
If you turned on sharing to the fediverse, you can see who liked your posts from other servers too. For the documented post-information paths, see Meta's help page on viewing post information. It also explains that the reply count includes all replies in the thread, not just direct replies to the original post.
Once the four numbers are in front of you, run them through our free Threads engagement rate calculator. It calculates both view-based and follower-based rates; choose the denominator that matches the comparison you want to make.

Work Through the Math on One Real Post
Time for a full worked example. The numbers are hypothetical, so swap in your own post once you see how the pieces fit. Say a post pulled 4,500 views, 120 likes, 20 replies and 10 reposts on an account sitting at 3,000 followers. Top line first: 120 + 20 + 10 = 150 interactions.
By views: (120 + 20 + 10) / 4500 x 100 = 150 / 4500 = 3.33%. By followers: 150 / 3000 x 100 = 5.00%. One post, two defensible numbers. Neither is wrong, they just answer different questions, which is exactly why you name the denominator.
Second example post, different shape: 900 views, 40 likes, 8 replies and 2 reposts on 6,000 followers. Top line: 40 + 8 + 2 = 50. By views: 50 / 900 x 100 = 5.56%. By followers: 50 / 6000 x 100 = 0.83%. Notice the shape of the numbers too: the view-based rate clears 5 percent because the post barely left your existing circle, while the follower-based rate sits under 1 percent on the bigger account. Small audiences flatter view-based rates and punish follower-based ones, so the same math reads differently at 600 followers than at 60,000.
Read the pair together. The view-based rate says the people who saw it responded well. The follower-based rate is low relative to the account's follower count, and it does not tell you how many followers actually saw the post. Two rates, two questions. Label which one you're quoting.
Quoting a rate without naming the denominator, view-based or follower-based.
Mixing the two denominators in one comparison and reading the gap as a trend.
Comparing a 2-hour post with a 2-week post instead of matching post age.
Reading the reply count as direct replies when it includes every reply in the whole thread.
Hunting for a universal good rate instead of beating your own trailing baseline.
Four Traps That Skew the Number
Trap one, inflated reply counts. On a thread you start, the reply count includes every reply in the whole thread, not only direct replies to your original post. If one reply sparks a side conversation, all of it lands in your top line. Know that before you celebrate a reply spike. The skew is real: a post with 6 direct replies and a 40-reply argument underneath shows 46 in the count. If you quote posts side by side, count the same way every time or the comparison quietly breaks.
Cross-platform views can change the denominator without a matching change in the interaction counts you're adding. Note that limitation before interpreting a view-based rate as a Threads-only response rate.
Trap three, mixed denominators. A 5.00% follower-based rate sitting beside a 3.33% view-based rate is not a trend, it's two different units in one sentence. Pick one denominator per comparison, name it, and hold it constant.
Treat the counts as platform-reported measures, not permanent ground truth: Meta labels likes, replies, reposts, quotes and views as in development. The labels are explained on the page about how Meta labels in-development metrics.

Compare Against Your Own Baseline
So is 3.33% good? Honest answer: there's no universal number, and this article won't invent one for you. Rates move with audience size, format and how views get counted. The only good that means anything is good against your own trailing baseline.
Build it like this: log each post's rate at 24 hours, same denominator every time, then average the last several posts. That trailing average is the line a new post has to beat. For the bigger question of what counts as good when benchmarks disagree, read what is a good engagement rate, our walkthrough of competing numbers and how to read them.
When one rate isn't enough context, our guide to Threads insights covers which parts of the insights panel deserve attention and which are vanity. Then, whenever a post beats your baseline, reread it before writing the next one. The lesson lives in what changed, not in the score.
Run the baseline math once and it stops being intimidating. Say your last five comparable posts at 24 hours scored 1.9, 2.4, 2.2, 2.6 and 2.9 percent by views. The median is 2.4. Today's post scored 3.1. That is above your recent middle, not a vague feeling of doing well, and it is the comparison your own audience actually set. Rebuild the baseline each week so recent reality, not last quarter's streak, is the bar.
A Weekly Routine That Stays Honest
Record each post's counts when it reaches the same age, such as 24 hours after publishing. Log the date, interactions, denominator and rate, then review the week's entries on one chosen day. A weekly check of current counts cannot reconstruct what every post showed at 24 hours, so capture them as posts mature. A notes app row, a spreadsheet column or a paper log all work. The format matters less than the discipline: every row needs date, post, age at measurement, interactions, denominator and rate.
Once a month, reread the log. Find the posts that beat the trailing average and the ones that missed, then change one variable at a time so the shift stays attributable. Cadence is one of those variables, and how often to post on Threads should be tested against your own numbers, not a magic figure from a stranger.
Keeping the method consistent makes changes easier to interpret, though a rate alone cannot establish what caused them.
Action checklist
Use this as the practical next pass after reading the guide.
- +Choose your denominator: views for post response, followers for cross-account comparisons.
- +Open your profile and post for the counts below it; use View activity to inspect who liked, reposted or quoted it.
- +Add the top line: likes + replies + reposts.
- +Divide by views or followers, multiply by 100, and write the rate down with its denominator.
- +Compare posts at the same age only, 24 hours as the default, against your trailing baseline.
- +Log every post in one place and reread the log monthly before changing anything.

Conclusion
That's the whole procedure: two formulas, where to find every count inside the app, four traps to dodge and a baseline habit that keeps the number honest.
Run it on tonight's post, log the result, and let your own trailing average be the judge that matters.



