TikTok carousels beat video on likes and lose to video on views, and both of those statements come from large studies. One found carousels earn 81% more engagement. Another found video gets five times the views. They are not contradicting each other. They are measuring different things on different accounts, and once you see which is which, choosing between a carousel and a video stops being a guess.
We sell TikTok views, likes and followers, so weigh the last section accordingly. The rest is about reading three studies properly, which costs nothing.
Three studies, three answers
Start with the numbers as published, because each one gets quoted alone as if it settled the question.
Fanpage Karma looked at nearly 700,000 posts from January to May 2025. On TikTok, carousels had an engagement rate more than 81% higher than video and almost 82% more likes. Reach was only 3% higher. Shares were about a third lower.
Metricool’s 2026 study of 2.3 million posts across 92,000 accounts found the opposite headline: video gets five times more views and six times more interactions than carousels.
Buffer sat in between. Its analysis of a million TikTok posts, reported by Music Ally, put video in front on views but described photo carousels and text posts as “surprisingly competitive”.
So carousels win by 81%, lose by a factor of five, or roughly draw. All three samples are big enough to take seriously. The disagreement is the useful part.
Why they disagree
What follows is our reading of the three, not something any of them states. Treat it as an argument.
A rate is not a total. Fanpage Karma’s headline is an engagement rate. Metricool’s is a count of views and interactions per post. A carousel can collect more likes for every person who sees it and still be seen by far fewer people. Both findings fit the same post.
The samples are different accounts. Each study is drawn from the customers of a different tool. A sample weighted towards established brand pages behaves differently from one weighted towards small creators, and none of the three is a random sample of TikTok.
A carousel is slower to leave. A viewer swiping through eight slides spends longer on the post than most videos hold them, with the like button under their thumb the whole time. That would lift likes without lifting reach, which is exactly the Fanpage Karma pattern: likes up 82%, reach up 3%.
The share number is the tell. Carousels were shared about a third less than videos. People like a carousel and keep it. They send a video. Sending is what carries a post to people who have never heard of you, and it is the difference between a format that holds an audience and one that finds a new one. We went through why those signals are not equal in TikTok saves vs likes.
Put together, the three studies describe one thing: a carousel does better with the people it reaches, and video reaches more people.
Everyone else has read the 81% too
There is a second number in Metricool’s study that explains why carousel advice is ageing badly.
Over the past year, the volume of video posts in its sample grew 72%. Images and carousels grew 140%. The format that was undersupplied in early 2025, when Fanpage Karma measured it, has since doubled and then some.
A format works best while few people are making it. The 81% was measured in a feed with far fewer carousels in it than the one you are posting into now. That does not make the figure wrong. It makes it a year and a half old, in a market that moved the whole time.
When a carousel is the right call
The evidence points to a narrow, clear use: content somebody wants to keep.
- The list. Seven tools, five mistakes, three places to eat. One item per slide. This is the native shape of the format.
- The steps. A recipe, a routine, a settings walkthrough. Viewers go at their own pace and come back to slide four, which a video makes awkward.
- The before and after. Two slides. The swipe is the reveal.
- The reference. Sizes, prices, dates, a checklist. Anything a person will screenshot is better as a slide than as a frame they have to pause on.
- The text-heavy point. If the video version would be you reading words off the screen, the words are the content. Let people read them.
The common thread is that the viewer controls the speed. Where that helps, make a carousel. Where it does not, you are giving up reach for nothing.
When video is the right call
Almost everywhere else, and for one reason above the others: reaching people who do not follow you.
- Anything that depends on a voice or a face. Stories, opinions, humour. Timing does not survive being cut into slides.
- Anything with motion in it. A demonstration, a transformation in progress, a reaction.
- Anything you want sent to a friend. The share gap is the widest and least flattering number carousels have.
- Anything meant to grow the account. Five times the views is not a margin you can design your way around with nicer slides.
Video also gives the feed a cleaner signal. Watch time on a video is a single well-understood curve. How TikTok scores a half-swiped carousel is not something it has published, and anyone who tells you the exact rule is guessing.
If you make a carousel, make it like this
A few things hold whichever study you believe.
- Slide one is the hook. It does the job of a video’s first second, and it has to do it as a still image. A title that promises something specific, large enough to read at a glance. Our guide to TikTok hooks applies word for word.
- One idea per slide. If a slide needs two reads, split it.
- Give a reason to swipe. Number the slides, or let a sentence run on to the next one. A slide that feels finished is where people leave.
- Stop when you are done. TikTok allows up to 35 images. No study we found shows an ideal count, so the honest rule is the boring one: as many as the content needs.
- Add a sound. A carousel still plays audio, and a silent post in a feed built on sound reads as broken.
- Write the caption for search. Slides full of text are still images, so the caption carries the words. That is covered in TikTok SEO.
The test that beats all three studies
Your account is not an average of 92,000 accounts. The only sample that describes it is your own.
Take one idea and make it twice: once as a video, once as a carousel. Post them a week apart at the same time of day. Do that three times with three ideas. Six posts will tell you more about your audience than any study, because it is the only test where the account and the topic stay the same and the format is the single thing that changes.
Compare views first, then saves and shares. Ignore likes on their own. They are the number carousels inflate most easily and the one that travels least, which is the point of working out your engagement rate properly.
Where buying views and likes fits
We sell TikTok views, so read this with the appropriate suspicion.
Bought views do the same thing to a carousel as to a video: they make it look seen. They do not make anyone swipe to slide two. If a carousel is weak on reach because the format is weak on reach, a bigger view count beside it changes how the post looks and nothing about how it spreads.
If you use them at all, use them on a post that is already being saved, and know that the platform removes engagement it judges inauthentic. That is set out in the risks of fake engagement.
The cheaper fix is the format decision itself. Picking the right container for the idea costs nothing and happens before you film anything.
Frequently asked questions
Do TikTok carousels get more views than videos?
On the largest study available, no. Metricool’s 2026 analysis of 2.3 million posts found video gets five times more views than carousels. Fanpage Karma found carousel reach 3% higher on an earlier and smaller sample. Neither supports the idea that carousels are a reach shortcut.
Do carousels get more engagement than videos?
More likes per viewer, yes. Fanpage Karma measured an engagement rate 81% higher and 82% more likes. The same study found about a third fewer shares, and shares are what carry a post to new people.
How many photos can a TikTok carousel have?
Up to 35. There is no reliable data on an ideal number. Use as many slides as the content has separate points and stop there.
Is TikTok pushing photo posts in the algorithm?
TikTok has not published anything that says so. What the data shows is supply: carousel volume grew 140% in a year in Metricool’s sample, against 72% for video. More carousels in the feed is evidence that creators moved, not that the feed favours them.
Should a small account post carousels or videos?
Mostly videos, because a small account needs reach more than it needs like rate. Add carousels for the things people keep: lists, steps and reference posts. Then run the same idea in both formats a few times and follow your own numbers.
Why did my carousel get likes but hardly any views?
That is the pattern the studies describe. A carousel holds the people who see it and earns likes from them, but it is shared less, so it travels less. A weak first slide makes it worse, because the first slide is the only one most of the feed ever sees.
The short version
The studies disagree because they measure different things. Fanpage Karma found carousels earn 81% more engagement and 82% more likes on nearly 700,000 posts. Metricool found video gets five times the views and six times the interactions on 2.3 million.
Both can be true. A carousel does better with the people it reaches, and video reaches more people. Reach was only 3% higher for carousels even in the study that favoured them, and shares were a third lower.
The advantage is also shrinking. Carousel volume grew 140% in a year, nearly twice the rate of video.
Use a carousel for content people keep: lists, steps, before and after, reference. Use video for everything meant to find a new audience.
Then test it on your own account. Same idea, both formats, three times. Six posts beat every benchmark, including the ones on this page.
