Likes are the easiest number to obsess over and the hardest to influence directly, because nobody decides to like a video. They decide to keep watching, and the like happens on the way out. Every tactic that works on likes is really a tactic that works on attention.
So this is not a list of tricks for the like button. It is a list of the things that get a video watched to the end by people who were not looking for you, because that is what produces likes at any scale worth counting.
What a like is worth to the algorithm
TikTok has described the For You system as ranking on user interactions, video information and device settings, with the interaction signals weighted by how much effort they take. A like is cheap to give, so it counts for less than a share or a full rewatch, and far less than watch time.
That ordering matters for what you do next. A like is best understood as a readout rather than a lever: it tells you the video landed, and it tells the next viewer the same thing. Nothing you do to chase likes on their own will move the distribution much, but everything that moves distribution shows up as likes.
The second job a like does has nothing to do with the algorithm. It is the first number a stranger reads. A video sitting at four likes says nobody else bothered, and people take that seriously in the half second before they decide to keep watching.
Work out your like rate first
Most people judge a video against the last one that did well, which is the worst available benchmark. Divide likes by views instead and you get a number you can compare across posts and against the rest of the platform.
Socialinsider’s 2026 benchmarks, built from two million videos across 214,507 profiles, put the platform-wide engagement rate at 3.85% of views in Q2 2026, counting likes, comments, shares and saves together. Likes are close to nine in ten of those interactions on a typical account, which puts an ordinary likes-only rate somewhere near 3 to 4% of views.
Use that as the frame. Under 2% means the video is being served and not landing. Around 3 to 4% is normal. Above 6% is genuinely strong and worth studying rather than repeating by accident. We go through the arithmetic, the tier benchmarks and the two ways people get this calculation wrong in what counts as a good TikTok engagement rate.
The first two seconds are the whole game
Nobody likes a video they scrolled past. The opening frame has to answer one question fast: is something going to happen here? Most videos that underperform are not bad videos, they are videos with the interesting part buried thirty seconds in.
- Start in the middle of the action. Cut the greeting, the setup and the throat clearing. If the first sentence could be deleted without losing information, delete it.
- Show the payoff before you explain it. The finished result, the reaction, the number, then the story of how you got there.
- Put text on the first frame. Sound is off more often than you think, and a line of text tells a scrolling viewer what they are looking at before the audio does.
- Do not promise what you cannot deliver. A hook that oversells produces a drop at second three, and that drop is the signal that ends the video’s run.
Watch time is the lever, likes are the readout
Completion rate is the number to protect. A forty second video watched to the end beats a three minute video abandoned at forty seconds, even though the watch time is identical, because the completion signal is what widens distribution.
The practical version: make the video as long as the idea, and no longer. Cut every pause. If a section exists so the video feels substantial, it is costing you the ending. Loops help for the same reason, since a video that reads as continuous gets rewatched, and a rewatch is worth more than a like.
Ask for the like, but ask for the right thing
“Like and follow” at the end of a video does almost nothing, because it arrives after the decision has been made and it asks for a favour. Asking a question works better, and it is measurable: Metricool’s 2026 study of 2.3 million posts found that posts with a question in the caption collected 26.19% more comments.
Comments are the point here, not politeness. A comment is a stronger signal than a like, it holds people on the video while they type, and comment threads pull viewers back for a second watch. Ask something specific that a stranger can answer in four words. Vague invitations to “let me know what you think” get nothing.
The small mechanical things that measurably help
None of these rescue weak content. All of them are free, and the same Metricool dataset gives a rough size for each.
- Hashtags. Posts using them saw nearly 5% more views and more than 9% more interactions. Hashtag-driven traffic also rose 114% year on year, which makes them worth the ten seconds they cost.
- Video over stills. Video posts pulled five times the reach and six times the interactions of image posts, even though image and carousel posting grew fastest.
- Timing. The busiest window is 6pm to 9pm with 8pm the peak hour. Posting into the window helps a little. Posting consistently helps more.
- Where the views come from. The For You feed supplies 72.70% of video views, so almost everything you are optimising for is being judged by people who have never seen your account.
Your video is finished in about ten days
Posts reach 96% of their total reach and nearly 98% of their total interactions within their first ten days. TikTok’s long tail is real but small, so a video that has not moved in two weeks is not going to.
That has a direct consequence for how you spend attention. Checking a three week old post is wasted effort. The compounding happens across posts, not inside one, which is why cadence beats polish for most accounts under 10,000 followers.
Expect lower numbers than last year, because everyone is getting them
Between early 2025 and early 2026, average video views per post fell 31.30% and interactions fell 31.17%, while the number of videos being posted rose 72.10%. More content is competing for a feed that did not grow at the same rate.
This is worth knowing before you conclude something is wrong with your account. A 30% decline in likes per video over the last year is the platform average, not a penalty. Judge yourself against your own recent posts and against the rate, not against the counts you were getting eighteen months ago.
When a video genuinely is not getting likes
There is a difference between a video that is not being shown and a video that is being shown and ignored. Views with no likes is the second case, and it is a content problem with a specific shape. Views near zero is the first case, and the fixes are unrelated. We separate the two, with the checks for each, in why your TikTok is not getting likes.
If a post is sitting at zero and you want it to stop looking abandoned while you work on the real fix, our free TikTok likes gives you 25 on a video you choose, once every 30 days, at no cost. Be clear about what that buys: it removes the empty-count problem and nothing else. Twenty five likes does not change distribution and does not make an ignored video interesting.
Buying likes: what it does and what it does not
We sell likes, so treat this section with the suspicion it deserves and check it against what we say on the rest of the site.
What a purchase does is change the social proof a stranger reads before deciding to watch. That effect is real and it is the only honest reason to do it. What it does not do is make TikTok show the video to more people. Bought likes are a weak signal on a metric that was already weak next to watch time, so if the video is not holding attention, a higher like count just means more people see an unloved-looking video with a bigger number on it.
The case where it makes sense is narrow: a video that is holding attention but looks under-liked next to its view count, usually in its first hours. If you want that, buy TikTok likes in a quantity that fits the video, which means keeping the resulting rate inside the bands above rather than buying a round number. Overshooting is the mistake that gets noticed, because a video with 4,000 views and 900 likes reads wrong to anyone who scrolls the comments.
Likes are not a payout
One correction worth making early, because it changes what you optimise for: TikTok does not pay for likes. Not per like, not at a threshold, not at a million. The monetisation programmes count followers, qualifying views and watch time, and the like count does not enter the calculation. We set out what actually pays and what the thresholds really are in how many likes you need to get paid on TikTok.
If your goal is money rather than reach, that reframes everything on this page. Likes remain a useful readout of whether a video landed. They are not the product.
The short version
- Likes follow watch time. Fix the hook and the ending before anything else.
- Judge videos on likes divided by views, not on the raw count. Around 3 to 4% is ordinary, above 6% is strong.
- Ask a specific question in the caption. Comments are worth more than likes and they produce likes.
- Hashtags, video format and the evening window each help a little. Cadence helps more.
- A post is done in ten days. Put the effort into the next one.
- Buying likes changes what a visitor concludes, not what the algorithm concludes.
If followers rather than likes are the thing you are actually short of, the mechanics are different enough to need their own guide: start with how to get more TikTok followers.
