The short version: The TikTok videos that get the most views aren’t the best-made ones, they’re the ones people watch all the way through or rewatch, and that single number, average watch time in seconds, matters more than likes, follows, or how many hashtags you cram into the caption. I’ve had a video filmed in a car outperform a studio-shot one by 340,000 views, and the reason is not luck.
The three numbers TikTok is watching
Forget the vanity metrics for a minute. TikTok’s recommendation system runs on three signals, and they’re not the ones on your dashboard front page.
- Average watch time, the number of seconds a video is watched, not the percentage. A 45-second video watched for 18 seconds beats a 12-second video watched for 6 seconds, even though both are 50 percent completion.
- Completion rate, whether people watch to the end. This matters more on short videos (under 15 seconds) than on longer ones.
- Rewatches and shares, which TikTok treats as a stronger vote than a like. A like costs nothing. A rewatch or a share costs attention, and TikTok knows the difference.
None of these are secret. TikTok has said as much in its own creator resources for years. What most people skip is what this means for the type of video you should be making, and that’s where it gets uncomfortable.
What this looks like on a real account
I posted a video in early 2025 about three AI tools I use for client proposals. Scripted, shot on a tripod, decent lighting, captions edited in, the kind of video you’d hand to a brand and they’d approve without a note. It got 41 views. Forty one.
Two days later I filmed myself in the car outside a client meeting, no script, talking about a client who’d ghosted me for three weeks after promising a six-figure retainer, and how I’d wasted half a day writing a proposal for nothing. Bad lighting. Wind noise. I didn’t even edit out the “um.” That one hit 380,000 views in 48 hours and is still, a year on, the single biggest driver of new followers on that account.
Same creator, same niche, same week almost. The difference wasn’t the topic, it was that the second video felt like something a friend would send you, not something a brand would approve. TikTok’s system rewarded the mess because the mess kept people watching.
The bit most advice avoids: polish is often the problem, not the fix
Every guide to TikTok growth tells you to improve your lighting, tighten your editing, add captions, use a hook template. All of that helps at the margins. But the uncomfortable truth is that a video that looks too finished often reads as an ad to both the viewer and, indirectly, to the algorithm, because ad-like videos get scrolled past faster and scrolling past faster tanks your watch time before the video ever gets a fair test.
I’ve watched small business owners spend three days editing a 30-second video with motion graphics and a licensed track, only to get beaten by a competitor’s 15-second phone clip with a typo in the caption. This isn’t an argument for being lazy. It’s an argument for spending your effort on the hook and the story, not the finish.
How long should the video be
There is no single correct length, but there’s a pattern worth knowing:
- Under 15 seconds: needs a completion rate near 50 percent or higher to get pushed past the first test pool of roughly 300 to 1,000 viewers.
- 30 to 60 seconds: TikTok cares more about total seconds watched than percentage, so a 45-second video watched for 20 seconds outperforms a 10-second video watched fully, in terms of raw watch time signal.
- Over 60 seconds: only works if the story earns it. A tutorial or a personal story with real stakes can run to two or three minutes and still get pushed hard, because long watch times per viewer are rare and valuable to the platform.
My personal rule after testing this across three accounts and roughly 400 videos: if I can’t justify every 5 seconds with a reason for someone to stay, I cut the video shorter. If the story needs 90 seconds, I let it run 90 seconds and don’t panic.
The first two seconds are doing 80 percent of the work
Nobody decides to keep watching your video at second ten. They decide in the first two seconds, sometimes the first frame. The videos that get the most views almost always do one of these things immediately:
- Open mid-sentence or mid-action, no intro, no “hey guys.”
- Put text on screen that creates a question (“the client who ghosted me just texted back” works far better than “story time”).
- Show the result first, then rewind to explain it.
The car video that hit 380k opened with me already mid-rant, no setup, just “so this client has done it again.” No hello. No context. That single choice probably accounts for a third of the views it got, because it stopped the scroll before the viewer’s brain had time to decide the video wasn’t for them.
Timing matters, but not as much as you’ve been told
Posting time affects the size of your first test pool, not whether the video eventually goes big. If you’re posting at 3am UK time and your audience is mostly in the UK, fewer real people see it in that crucial first hour and the algorithm has less data to work with. I’ve written a full breakdown of the best day and time to post on TikTok in 2026, but the short version is this: post when your specific audience is awake and scrolling, test a few slots for two weeks, and stop treating timing as the reason a video flopped. I’ve had videos posted at 2pm on a Tuesday outperform ones posted at the “optimal” 7pm slot by ten times, because the content, not the clock, did the work.
Your follower count barely decides who sees you
This trips up more people than anything else on this list. TikTok shows most videos to non-followers first. An account with 300 followers can outperform an account with 300,000 followers on a single video, because the platform tests every video fresh against a small sample and only pushes it wider if that sample engages. I’ve covered the real thresholds around growth and monetisation in how many TikTok followers you need before you can earn money, and one thing that surprises people every time is how weak the correlation is between follower count and view count on any given video. Follower count decides your baseline, not your ceiling.
Comments and shares do more than likes ever will
A video with a modest number of likes but a comment section people are reading and replying in will often keep climbing for days, sometimes weeks, because TikTok reads sustained comment activity as a sign the video still has legs. This is also why it’s worth knowing why TikTok comments get folded and how to stop it happening to you, because a folded comment section quietly kills the exact signal that would have kept your video circulating.
Shares matter even more than comments, because a share moves your video off TikTok entirely, into a DM or a WhatsApp group, and brings it back with a fresh viewer attached. If you want to understand how much reach a single share can generate, it’s worth reading what TikTok’s analytics reveal about who shared your video. The videos that get the most views are, almost without exception, the ones people send to a specific person rather than just double tap on.
A quick audit you can run this week
If you want to know why your last ten videos performed the way they did, do this:
- Open your TikTok analytics and note the average watch time (in seconds, not percentage) on your five best and five worst videos.
- Check where the drop-off happens on your worst videos. If it’s in the first three seconds, your hook is the problem, not the topic.
- Count how many videos opened with a greeting, an intro, or a logo before getting to the point. Cut those out on the next five and compare.
- Look at your comment count relative to your view count on your best video. Anything above roughly 1 comment per 500 views is strong and worth studying for what prompted people to type something.
- Post the same core idea twice, once polished and once rough, a week apart, and compare watch time, not likes. Most people who do this are surprised by the result.
This takes about 40 minutes and tells you more than any generic “post consistently” advice ever will.
Related reading: tik tok videos keep freezing.
Related reading: 300 views in 30 minutes tiktok.
Related reading: best songs on tiktok.
Frequently asked questions
Do hashtags affect how many views a TikTok video gets?
Barely, at this point. Hashtags help TikTok categorise your video slightly but they are not the deciding factor in reach; watch time, completion rate, and shares carry far more weight, and one or two relevant tags do the same job as fifteen stuffed into the caption.
Why did my well-edited TikTok video get fewer views than a rough one?
Polished videos often read as ads and get scrolled past faster, which lowers watch time before the video has had a fair test. A rougher, more personal video frequently keeps people watching longer, which is the exact signal TikTok rewards with wider distribution.
Does posting every day increase views on TikTok?
Not automatically. TikTok compares each new video’s early performance to your own recent average, so a run of weak videos can make the algorithm treat your next video more cautiously. Consistency helps you learn faster, not because volume alone triggers reach.
How long should a TikTok video be to get the most views?
There’s no single right length. Videos under 15 seconds need a high completion rate to spread, while longer videos over 60 seconds rely more on total seconds watched. The safest rule is to keep the video only as long as the story or point needs, then cut.
