Most answers to this question hand you one percentage to aim for. X publishes no organic completion benchmark, so a good Twitter video completion rate is the one that beats your own videos of the same length and purpose.
What counts as a good video completion rate on Twitter?
A good Twitter video completion rate is one that holds up against your own videos of similar length, topic and purpose, measured as the share of video starts that reach 100% Watched. Circleboom measures every native X video stage by stage, from Video Started through 25%, 50%, 75% and 100% Watched, with data taken from X's Enterprise-level API.
→ Twitter video completion rate
Why one benchmark for Twitter video completion fails
A universal target fails because completion depends on three things that change from video to video: how long the clip runs, what it is for, and how many of its starts were real viewing. A number that ignores all three describes nobody's account.
Search results for this question tend to offer rules of thumb, such as a rate above one figure being "solid" and above another being "exceptional." None of them cite a source.
Ad-side figures exist too, but they describe paid placements, 15-second view bidding and relative lifts, which tell an organic creator very little about their own timeline.
Length is the biggest distortion. Finishing a 20-second clip and finishing a six-minute tutorial are different amounts of attention, so a single target rewards short videos and punishes long ones regardless of quality.
Purpose is the second. A teaser exists to send people somewhere else. A tutorial exists to be watched through. Holding both to one bar misreads at least one of them.
The third distortion is the base. Starts include timeline autoplay, so two videos with identical viewing behavior can report different rates simply because one autoplayed in busier feeds.
Why the wrong benchmark costs more than one post
This matters beyond a single post. Creators use completion to decide where production effort goes next: which explanations to expand, which formats to drop, which series to keep.
A wrong benchmark sends that effort to the wrong place, and the cost shows up weeks later as a content plan built on a misread.
The practical fix is to replace the borrowed number with your own baseline. The X video completion rate for each native upload, read next to clips of the same length band, gives you that baseline without guessing. Circleboom shows those stage counts for every eligible X video in a date range, so the comparison set builds itself as you post.
How is Twitter video completion rate calculated?
Twitter video completion rate is the number of viewers who reached the 100% Watched stage divided by the number of video starts, shown as a percentage. A video with 800 starts and 120 completions has a 15% completion rate.
The calculation is simple. The denominator is where the arguments start.
X's API exposes the underlying stage counts as playback quartile metrics: counts at 0%, 25%, 50%, 75% and 100% of the video, listed among X's non-public metrics. Those quartiles are the raw material for any stage chart, including Circleboom's.
Views follow a different rule. X's note on common analytics discrepancies explains that quartile and completion counts do not require the in-view condition a counted view does. So completions and views are measured differently, and dividing completions by views mixes two counting systems.
That mix produces numbers that look precise and mean little. A video completion rate on Twitter calculated against views can rise or fall with the view rule rather than with how people watched, and it cannot be compared with a rate calculated against starts.
Pick starts as the base and keep it for every video you compare.
Share of starts versus stage-to-stage retention
The same stage counts support two different readings, and mixing them up is a common error. Take that 800-start video further, as an illustrative example:
| Stage | Count | Share of starts | Kept from the previous stage |
|---|---|---|---|
| Video Started | 800 | 100% | not applicable |
| 25% Watched | 360 | 45% | 45% |
| 50% Watched | 240 | 30% | 67% |
| 75% Watched | 170 | 21% | 71% |
| 100% Watched | 120 | 15% | 71% |
The share-of-starts column gives the completion rate and lets you compare videos. The stage-to-stage column shows where viewing thinned fastest. Here, the first quarter lost the most, while each later quarter kept most of the viewers who arrived.
Divide completions by starts, and write the base down every time you quote the rate. A percentage without its denominator cannot be compared with anything, including your own last video.
A framework for benchmarking X video completion against your own videos
Benchmarking against yourself means grouping videos that should behave alike, then comparing the share of starts each one keeps at every stage. The table below is the framework I use whenever someone asks me whether a number is good. It needs no outside figure at all.
| Comparison axis | What to hold constant | What to compare |
|---|---|---|
| Length band | Videos of similar runtime, such as under 30 seconds, 30 to 90 seconds, or several minutes | Share of starts reaching 100% Watched |
| Purpose | Teasers with teasers, tutorials with tutorials, announcements with announcements | Whether viewing reaches the part that serves the goal |
| Topic | Same subject family or series | Which explanation or format holds people longer |
| Stage shares | The same denominator (Video Started) for every stage | Share at 25%, 50%, 75% and 100%, not raw counts |
| Exposure | Similar distribution, read from impressions and starts | Whether a low result is weak retention or limited reach |
How many videos make a fair comparison set
One earlier video is a data point, not a baseline.
Use repeated relevant examples before deciding which explanation or format merits more work. Several clips in the same length band and with the same goal give a range, and a new video lands inside it, above it or below it.
Keep exposure in view while you build the set. A video that reached fewer people can still show strong stage shares, so separate limited exposure from weaker retention among the starts you did get.
Fill one row per video, month after month, and your Twitter video completion rate stops being a lone number. It becomes a position inside a range you built yourself.
Raw counts mislead on their own. A video with more starts can show more completions while keeping a smaller share of its starting audience, which is why every cell above compares shares of starts.
Reading the framework with a 15% and a 40% example
Take two videos from one account. A 2-minute video keeps 15% of its starts to 100% Watched. A 45-second video keeps 40%.
Read against a universal benchmark, the first looks weak and the second looks strong. Read against the framework, they sit in different length bands and should not be ranked against each other at all.
What the pair does suggest is that shorter cuts are worth testing for that audience.
The verdict comes from the comparison set, not from the percentage. A 15% rate is good if your other 2-minute tutorials sit at 9%. It is a warning if they sit at 25%.
What a good result looks like by video purpose
The stage that matters most shifts with the job the video was made for:
- Teaser or trailer: the opening quarter carries the message, so the 25% share matters more than the end.
- Tutorial or walkthrough: the late stages matter, because the payoff usually sits near the end.
- Announcement: the first half matters, since the news should land before viewers drift.
- Conversation starter: completion matters less than the replies that follow.
None of these comes with a fixed percentage. Each tells you which bar in the chart to compare against your own earlier videos of the same kind.
A video that clearly beats its own band is worth more than one post. Treat it like an evergreen tweet and give it another turn in front of your audience.
Where Circleboom fits in judging X video completion
Circleboom gives you the stage counts the framework needs. Its Video Analytics feature reads each natively uploaded X video and draws a horizontal bar chart of Video Started, 25%, 50%, 75% and 100% Watched, so the drop between stages is visible at a glance. To reach it, open Circleboom Twitter and go to the post analytics section for the connected account.
As a verified Enterprise partner of X, Circleboom pulls those numbers through the official route rather than scraping. The chart shows the counts X's API returns for each post. When you compare your own videos against each other, accurate base data is the whole point.
You can set a date range and review each video's retention chart in turn to compare them. The Twitter video completion rate inputs for each one are there without paying for X Premium.
When I review a month of videos, I sort them by length band first and only then look at the 100% bar. That order stops a short clip's high X video completion rate from making a longer tutorial look like a failure.
Limits that shape the comparison
Three limits shape what you can benchmark:
- Only video uploaded natively to X at posting time produces stage data; posts that link out to YouTube or Vimeo are excluded.
- Older posts may return partial stages when X's API no longer provides granular retention for them.
- Counts reflect the moment of retrieval, so refresh before comparing a fresh video with an older one.
The view is read-only. Decisions about length, hook or posting time are acted on in X Post Planner, not from the chart.
A related walkthrough on the reach side of the same posts: how to read the impressions and engagements behind your tweets, the two numbers that sit on either side of a completion rate.
Should you trust completion rate or engagement rate on Twitter?
Neither is better in general; they answer different questions. Completion rate shows how far viewing continued through the video. Engagement rate shows how many people responded with likes, reposts or replies. The better signal is the one that matches what the video was made to do.
A tutorial or explainer can succeed quietly. Good completion with few replies still fits an informational purpose, so silence is not proof of disinterest.
A video built to start a conversation, on the other hand, should be judged on replies and reposts even if many viewers drop off early.
How the engagement side is judged has its own logic. The breakdown of what is a good engagement rate on Twitter X covers that half, and it pairs naturally with a completion read.
In practice I read them together. When completion is strong but engagement is weak, I compare the same posts in Circleboom's engagement analytics before deciding the video needs to change.
If a quick per-post figure is enough, the Twitter tweet engagement calculator for free gives one without a full report.
Which X video metrics to read next to completion rate
Read completion rate next to impressions, starts, the stage shares at 25%, 50% and 75%, and post engagement. Each one answers a question completion cannot answer alone.
- Impressions: how many times the post was shown, which separates weak distribution from weak retention.
- Video starts: the base of the completion rate, including timeline autoplay.
- Stage shares: where viewing thinned, by quarter of the video.
- Post engagement: whether watching turned into likes, reposts or replies.
Low distribution and weak progression through the video need different investigations. If stage shares look healthy but impressions are thin, the work is distribution. If starts are plentiful and the 25% bar collapses, the work is the hook.
For a wider view of what belongs in a regular report, see the key Twitter performance metrics you should track and why they matter.
Measurement habits matter as much as the metric list. The advice on how to measure your Twitter engagement and be more social suits checking the same numbers regularly, not only when a video disappoints.
A simple monthly log covers it. For each new video, note the Twitter video completion rate with its base, the impressions, the share at 25%, and the engagement count.
Four columns are enough to separate a distribution issue from a hook issue and from a quiet but well-watched success.
All of these sit in one place in Circleboom's Twitter post analytics, where video stages and post engagement are read for the same posts.
The only completion benchmark that holds on X
A good Twitter video completion rate is the one that beats your own videos of the same length and purpose, counted from starts, every time.
→ Check the completion rate of each Twitter video
What people ask about video completion on X
Do autoplay views count as video starts on Twitter?
Yes. Start counts include autoplay views from the X timeline, so a video that begins playing while someone scrolls past registers a start without deliberate viewing. That pulls the completion rate down, which is why comparing starts with likes, reposts and replies helps gauge real interest.
How do I know which part of my X video to fix?
Look for the largest drop between two stages in Circleboom's Video Analytics. A big loss before 25% points at the opening quarter, while a larger drop between later stages marks that interval for review. The chart shows where to look, not the exact second or the reason viewers left.
){kind=link}