TikTok analytics for business comes down to seven numbers: views, retention rate, average watch time, watched full video, shares, profile views and new followers, each read against your own account’s median rather than against viral accounts. TikTok reports them in TikTok Studio’s analytics and, for Verified Business Accounts, in the Business Suite, while ad results live in TikTok Ads Manager. The rule for making more: one post at 2x your median is a lead worth remaking, and a format has earned new versions when at least three of its posts beat your median and those wins are more than half of its posts.
Key takeaways
- Read retention first. TikTok suggests watching it first, ranks engagement second and says watch time is more heavily weighted (TikTok Creator Academy, updated September 2026).
- Judge each post at 7 days against your own median. TikTok says analytics can take 1 to 2 days to update.
- One post at 2x your median is a lead. A format where at least three posts, and more than half of them, beat your median is a pattern.
- Keep organic and paid apart: TikTok attributes the views, likes and follows a Spark Ad gains to the original post (TikTok Ads Help Center).
- No business account needed: TikTok Studio analytics are free and open once you have one public post.
Which TikTok metrics show a brand post is working?
Seven numbers, read in order: did people stop, did they stay, did they act.
| Metric | Where to find it | What it tells you | How to judge it against your own baseline |
|---|---|---|---|
| Views | Overview tab, each post, Business Suite Video tab | Whether TikTok kept showing the post | 7-day views against the median of your last 15 to 20 posts |
| Retention rate | Each post’s analytics | How much of the video people watched, and where they dropped off | Share still watching at 2 to 3 seconds, compared across posts |
| Average watch time | Each post’s analytics | How long the average viewer stayed | As a share of video length, only against posts of similar length |
| Watched full video | Each post’s analytics | Finish rate: the share who reached the last few seconds | Within the same length band, since it moves with video length |
| Shares | Overview tab and each post | Whether viewers thought it worth sending on | Shares per 1,000 views against your median |
| Profile views | Overview tab | Whether a post made people curious about the brand | Profile views in the week a post went out against a usual week |
| New followers per post | Content tab (top posts by new followers) | Whether views turned into an audience | Follows per 1,000 views, by format |
Where each number lives comes from TikTok’s Introducing TikTok analytics (updated October 6, 2026), which also defines watched full video as the share of people who watched until the last few seconds and notes it fluctuates with the video’s length.
Two cautions. Views are not people: for unique viewers, use the Viewers tab, which is the closer measure of organic reach. And the age and gender figures on that tab come from TikTok’s predictive models rather than confirmed user data, so treat them as estimates.
Where do you find TikTok analytics for a business account?
You don’t need a business account. TikTok’s account types page lists analytics in TikTok Studio for General accounts and in the Business Suite for Verified Business Accounts, and both need at least one public post. TikTok calls its analytics a free performance-tracking tool.
- TikTok Studio. In the TikTok app, go to your profile, tap the menu, then TikTok Studio and Analytics. It is also the Analytics tab in TikTok Studio on the web and in the TikTok Studio app. Pick the last 7, 28, 60 or 365 days or a custom range, and download up to 365 days of data at a time (TikTok Creator Academy, October 2026).
- Business Suite. Verified Business Accounts and accounts with Advanced Access get the Business Suite, in the TikTok app or on the web. On the web, its Analytics tab splits into Overview, Audience and Video, which sorts your videos by newest or most viewed (TikTok Ads Help Center, updated May 2026).
Three quirks trip brands up, all from TikTok’s My posts aren’t getting views help page as of October 2026:
- Analytics are only generated for videos posted after you enable the feature.
- Data can take 1 to 2 days to update, so a day-one read undercounts.
- Follower activity shows in your local time zone, but all other analytics are in UTC, so a 9 p.m. post in New York lands on the next day’s row.
How do you judge a post against your own baseline?
Use your own account as the benchmark: TikTok’s Creator Academy calls fluctuations between videos normal. Averages mislead because views are heavily skewed: most posts get modest numbers and a few go viral. In Buffer’s study of 11.4 million TikTok posts (October 2025), median views per post stayed between 459 and 506 at every posting frequency, while the 90th percentile, the line into the top 10% of posts, sat at 3,722 views for one post a week and 14,401 for 11 or more posts a week. A median shrugs off those spikes; an average doesn’t.
The same Creator Academy article says follower count and previous high-performing videos are not direct factors in TikTok’s recommendation system, one more reason to judge each post on its own numbers.
- Pick the metric that matches the job. Retention for a hook, watch time for the body, follows for audience building, views for reach.
- Log every post at 7 days. Record each post’s numbers in a sheet on its 7th day, after the 1 to 2 day lag has cleared, so every post is judged on the same window.
- Set the baseline. The median 7-day value of your last 15 to 20 posts. Reset it monthly.
- Tag each post. Format (list slideshow, demo, talking head, myth vs fact), hook angle and length.
- Compare like with like. A 9-second slideshow and a 50-second talking head don’t share a watch-time baseline.
- Leave out boosted posts. A Spark Ad’s views, likes and follows count toward the organic post (Ads Help Center, updated June 2026), so a boosted post is no longer an organic read.
What is a good TikTok engagement rate for a brand?
It depends on what you divide by. Quid’s 2026 benchmark puts the median brand at 2.01% per TikTok post, measured as interactions divided by followers. Socialinsider divides likes, comments, shares and saves by views, and found an average of 3.85% in Q2 2026. Pick one formula, use it every time, and compare each post with your own median. More benchmarks are in TikTok marketing statistics.
When should you make more posts like a winner?
Our working rule of thumb: a format earns more versions when at least three of its posts beat your median on the metric you care about, and those wins are more than half of its posts. The second half of the rule matters because about half of your posts land above your median by definition. A single post at 2x your median or more is a lead worth remaking, not proof, because one post can take off for reasons unrelated to the format.
“More like it” means changing one thing at a time, which is the core of creative testing:
- Same format, new hook.
- Same hook, new format.
- Same idea, new opening frame.
Here is the rule on a hypothetical account (made-up numbers, not a Relay customer). Say a coffee brand’s median 7-day views across its last 18 posts is 1,100. Four list slideshows came in at 2,400, 1,600, 1,300 and 700. Five founder talking heads came in at 3,900, 800, 600, 500 and 450. The talking head has the single biggest post, but only one of five beat the median. The slideshows beat it three times out of four (75%). So the next batch is slideshows with new hooks, plus two remakes of the 3,900-view talking head, which is about 3.5x the median, before anyone calls it a format.
What do common patterns usually mean?
| What you see | What it usually means | What to change next |
|---|---|---|
| Most viewers gone by 00:02 | The opening didn’t earn attention (a missing hook is the first thing TikTok tells creators to check) | Same body, new hook |
| Strong early retention, low watch time | The hook promised something the middle didn’t deliver | Keep the hook, tighten the body |
| Good views, few follows | People liked the post, not the account | Turn it into a series and make the profile line clearer |
| A growing share of views from Search | People find you by query | Reuse those search terms in future titles, as TikTok suggests |
The first and last rows follow TikTok’s own advice in Using analytics as a tool to improve video performance. The middle rows are what we act on after a lot of TikTok tests.
How are TikTok ad analytics different from organic analytics?
Organic and paid are measured in different tools, and mixing them blurs both reads.
Organic posts: TikTok Studio. Everything above. The data is per post, but noisy: posts go out on different days to different viewers, so look for wide gaps and repeat winners.
Paid ads: TikTok Ads Manager.
- Reporting. Video play metrics include 2-second and 6-second video views, views at 25%, 50%, 75% and 100% of length, and average play time per video view (TikTok Ads Help Center, updated November 2025). Divide 2-second video views by impressions to get hook rate.
- Video Insights. It shows each video’s CTR, total cost, CPC and CPM, a Key Frame Analysis graph that plots two metrics you choose (clicks and video views, for example) second by second, and Industry Benchmark rankings (TikTok Ads Help Center, updated February 2025).
- Split Testing. It divides your audience into two equal groups, shows each one version, and names a winner only when the result is statistically significant at a 90% confidence rate (TikTok Ads Help Center, updated January 2026).
Other brands’ ads: Creative Center Top Ads. The Top Ads dashboard (updated May 2025) lets you filter high-performing ads by region, industry and campaign objective and sort them by reach, CTR or view rate. It is research, not your data.
| Question | Organic posts | Paid ads |
|---|---|---|
| Did the opening work? | Retention at 2 to 3 seconds | 2-second video views ÷ impressions |
| Did it hold? | Average watch time, watched full video | 6-second views, views at 25% to 100%, average play time per video view |
| Did people act? | Shares, profile views, new followers | CTR, CPC |
| Clean A/B comparison? | No: read wide gaps and repeat winners | Yes, with Split Testing |
The bridge between the two is the post itself: test hooks organically, then put money behind proven posts as Spark Ads. Step by step: how to turn organic TikToks into Spark Ads.
Where does Relay fit?
Reading analytics is the easy half. The hard half is having enough posts to read and remaking the winners, which is the part Relay does.
Relay looks at what is getting views in your niche, then makes slideshows and videos for your brand in your voice. You swipe to approve, editing the words first if you want. Connect your TikTok once and Relay sends each approved post to your TikTok inbox on its day. You see every post’s views in one place, and when one does well, Relay makes new takes of it, so you can compare hooks as performance data comes in. That covers the remake step without writing every version yourself; checking a format against your median is still a TikTok Studio job.
For a 90-day schedule built on this loop, see how to grow a brand TikTok from zero.
When is Relay the wrong tool for TikTok analytics?
- You want a full analytics suite. Relay is not an analytics tool: it shows every post’s views so you know what to remake. For retention curves, watch time, traffic sources and follower data, use TikTok Studio, whose analytics TikTok describes as free.
- You report across Instagram, YouTube and TikTok. Relay’s self-serve product works with TikTok only, so a cross-platform tool fits better (see the best TikTok scheduling tools).
- You only run paid ads. Self-serve Relay doesn’t manage ad spend; Ads Manager reporting, Video Insights and Split Testing are the tools for reading paid results. If what you need is more ad creative, Relay’s managed UGC ads service is the closer match.
- You already make more posts than you can review. Your bottleneck is analysis, and a weekly check against your median in TikTok Studio may be all you need.
Start with your own median
Pull your last 20 posts from TikTok Studio, find the median 7-day views, tag each post by format and see which format beat it at least three times and more often than not. If you need more posts to test than you can make, Relay makes slideshows and videos for you to approve, then makes new takes of the ones that do well. Or start with testing hooks before paid ads.