Carousel or Reel? The format studies only disagree until you check what each one divides by

KEY TAKEAWAYS

1. Reels reach about 1.8 to 1.9 times as many accounts as carousels in Fanpage Karma's 2026 data, and two to three times as many non-followers.
2. Carousels earn about twice the engagement per person reached, but total interactions per post are close to a tie.
3. Views and reach are different counts. Carousels lead on views because repeat displays count.
4. Pick the format from the job of the post, then check reach per follower, interactions per reach and views per reach on your own account.

Instagram put a carousel editor into its Edits app this week. The launch was reported on October 8, 2026, for iOS first, with Android to follow. OnlineMarketing.de had spotted it as a limited test on September 4. Dan Moreno, a design engineer on the Edits team, said that when creators were asked what the app should do next, carousels "came up again and again."

So carousels are about to get easier to make, and the old question returns: carousel or Reel? The published data seems to give opposite answers. It does not. Each study counts a different thing and divides by a different base. This article lines them up, shows what they agree on, and turns that into a choice you can make post by post.

Seven datasets, three denominators

Here is what the main sources report. The ratios in the last column are our arithmetic.

Read down the last column and it looks like a fight. Read the second column and it stops being one. Three of these rows count people reached. Two count interactions against different bases. One counts interactions with no base. One counts views.

Why the doubled engagement mostly disappears

Buffer's 2026 figure is the one most guides quote: carousels at 6.9%, Reels at 3.3%, which Buffer describes as 109% more engagement per person reached. That is a true statement about a rate. It is not a statement about how many interactions a post collects.

A rate divided by reach rises when reach falls. Carousels reach fewer people, and a larger share of those people already follow the account. A likely result is that the per-person rate doubles while the total barely moves.

The other rows confirm it. Divide by followers, as Socialinsider does, and the gap in the second quarter of 2026 is 0.50% against 0.48%. Buffer's own 2024 analysis put carousels at 1.12 times the engagement of Reels in absolute terms. Metricool's averages run the other way, with Reels about 15% ahead. Averages are pulled up by a few viral videos, so that direction is expected. SocialCrawl's sample of public video view counts in August 2026 had a mean about 40 times its median.

One caveat on Buffer. The format article states the Instagram rate as a share of reach. The landing page for the underlying report gives a formula that divides by followers. We have followed the article's wording, but treat the 6.9% and 3.3% as indicative.

What the evidence supports: per post, total interactions are close to a tie. The formats differ in who sees the post and what those people do.

Views are not reach

Socialinsider's views table shows carousels ahead in every size band. At 100,000 to 1 million followers it is 35,370 views against 16,035. Fanpage Karma shows Reels reaching nearly twice as many accounts. Both can hold.

SocialCrawl's August 2026 review quotes Meta's help pages on the definitions. A Reel view is logged each time the video starts to play or replays. A view of a feed post or carousel is logged each time it is displayed. Accounts reached counts unique accounts. So views include repeats, and reach does not.

Carousels collect repeats for a known reason. Adam Mosseri, head of Instagram, said in remarks reported in October 2024 that a carousel someone saw without swiping is often shown to them again. Pulse Advertising and RouteNote both reported the remark. Note what he was comparing. His point was why carousels out-reach single photos. Several guides now present it as an advantage over Reels, which he did not claim.

This also resolves the loudest contradiction. Net Influencer's October 2025 write-up of the Metricool and HypeAuditor playbook, built on about 700 million posts from January to June 2025, gives brand accounts over 1 million followers a median of 217,668 for carousels against 110,500 for Reels. The article uses reach and views as the same word. Metricool's own press release for that study says Reels lead in reach across follower tiers. The carousel figure is most likely views. On unique reach, the 2026 data has Reels ahead at every size.

Meta's developer documentation adds a reason for humility. It labels views as a metric in development and reach as estimated, and it returns no insights for the individual items inside a carousel. Per-slide claims from third-party tools rest on something other than Meta's API.

A worked example with invented numbers

This account is hypothetical. It has 8,000 followers and behaves exactly like the medians, which no real account does.

A Reel reaches 4.9% of followers, or 392 accounts. A carousel reaches 2.6%, or 208. Apply Buffer's per-reach rates, which come from a different dataset. The Reel earns 3.3% of 392, about 13 interactions. The carousel earns 6.9% of 208, about 14.

Same total. Different audience. Fanpage Karma found that under 10,000 followers, Reels reach nearly three times as many non-followers as carousels. So the Reel's 13 interactions include more strangers, and the carousel's 14 come mostly from people who already follow.

That is the real choice. It is not which format scores higher. It is which group the post is for.

Which format for which job?

Two rows deserve a note. Saves are the claim carousels are best known for, and Metricool does report nine times the saves of single images. Against Reels, the median gap is two saves. And the carousel case does not travel. Metricool found Instagram carousels get 4.7 times the views and 5.8 times the interactions of TikTok carousels.

A seven-step method for your own account

1. Write the job before the format. One line per post: who it is for and what you want them to do. Use the table to pick a starting format.
2. Export 20 to 30 recent posts per format. For each, record accounts reached, the follower and non-follower split, views, likes, comments, saves, shares and follows.
3. Compute three ratios per post. Reach divided by followers. Interactions divided by reach. Views divided by reach, which shows repeat exposure. Keep the denominators labelled.
4. Add the two ratios Instagram has named. Mosseri said in January 2025 that the top ranking signals are watch time, likes and sends, and pointed creators to likes per reach and sends per reach. He added that likes weigh slightly more with followers and sends slightly more with non-followers.
5. Use medians, not averages. One viral Reel will make the format look better than it usually is.
6. Run paired posts. Take one idea and publish it as a Reel and as a carousel, a week apart, alternating which goes first. Hootsuite did this with six pairs on one account in 2023, and carousels took slightly more reach there, 2,316 against 2,052 in total. That is the opposite of the large datasets, and it shows why your own pairs matter. Six pairs give a direction, not a proof. The same caution about small samples applies as in testing ad concepts.
7. Repeat each quarter. Supply is shifting. Metricool recorded Reels publishing up 35% and carousels up 24% year on year in early 2026, and a carousel editor in Edits will add more.

For Reels, read the opening separately. The metrics for that are in our piece on the 3-second hook rule. If a creator is involved, whose handle publishes is a separate decision from format.

What the popular guides leave out

We read the pages that rank for this question. They agree on a slogan: Reels for reach, carousels for engagement. The slogan is roughly right and hides three things.

First, none of them state the denominator. PostEverywhere quotes a carousel engagement rate of 1.92% against 0.50% for Reels with no base given. Second, none work out that a doubled rate on half the reach is the same number of interactions. Third, several give precise swipe rules. TryMyPost says a carousel is pushed to Explore when 60% of viewers reach the last slide. It cites no dataset, and Meta's API exposes no per-slide insights to check it against.

Where this stops

Every dataset here comes from a company that sells social media software, and each is drawn from that company's users. All are observational. Fanpage Karma says plainly that its results do not prove format caused the difference. People choose a carousel for one kind of idea and a Reel for another, so the formats are never carrying the same content.

Fanpage Karma does not disclose how many accounts sit behind its 3 million posts. Socialinsider's rate counts only likes and comments. Metricool's per-post figures reached us through a trade publication, not the full report.

None of that changes the practical reading. The studies do not disagree about what happens. They disagree about what to divide by. Decide who the post is for, pick the format that reaches them, and measure with the same three ratios every time.

If you want the same kind of tracing applied to how your brand shows up in search and AI answers, the waitlist is open.

Questions

Do carousels get more reach than Reels on Instagram?

Not in the largest recent dataset. Fanpage Karma's analysis of more than 3 million posts from January to April 2026 found Reels ahead in every follower band, at 4.9% of followers against 2.6% for carousels under 10,000 followers. Carousels often show more views, because views count repeat displays.

Why do some studies say carousels get double the engagement of Reels?

Because they divide by reach. Buffer reports a median of 6.9% for carousels and 3.3% for Reels as a share of people reached. Carousels reach fewer people, so the rate rises. Divided by followers, Socialinsider found 0.50% against 0.48% in the second quarter of 2026. Per post, the totals are close.

Does Instagram show a carousel twice?

Adam Mosseri has said a carousel that someone saw but did not swipe is often shown to them again. He gave that as a reason carousels out-reach single photos. He was not comparing them with Reels. It does help explain why carousels record more views than unique accounts reached would suggest.

Should I use carousels on TikTok as well?

The evidence points the other way. Buffer's 2026 data gives TikTok video a median engagement rate of 3.39% against 1.92% for carousels and photo posts. Metricool reports that Instagram carousels get 4.7 times the views of TikTok carousels. Treat the carousel case as specific to Instagram.

Sources

1. Net Influencer, Instagram Adds Carousel Editor To Its Edits App netinfluencer.com
2. Social Samosa, Instagram adds carousel creation tools to Edits app socialsamosa.com
3. OnlineMarketing.de, Edits testet Carousel Editor onlinemarketing.de
4. Fanpage Karma, Which Instagram format gets the highest reach? fanpagekarma.com
5. Buffer, Best Content Format on Social Platforms in 2026 buffer.com
6. Buffer, Data Shows Instagram Reels are Best For Reach, But Not Engagement buffer.com
7. Buffer, The State of Social Engagement in 2026 buffer.com
8. Socialinsider, 2026 Instagram Organic Benchmarks socialinsider.io
9. Socialinsider, Instagram Engagement Report socialinsider.io
10. Metricool, 2026 Instagram Study press release metricool.com
11. Net Influencer, Instagram views up 27 percent in early 2026 as publishing increases netinfluencer.com
12. Net Influencer, Instagram carousels outperform Reels for larger accounts netinfluencer.com
13. Metricool, Metricool and HypeAuditor Unveil 2025 Instagram Content Playbook metricool.com
14. SocialCrawl, Instagram Reels Views, What Counts (We Measured 425 Posts) socialcrawl.dev
15. Meta for Developers, Instagram Media Insights developers.facebook.com
16. Pulse Advertising, Boost Instagram Reach with Carousels pulse-advertising.com
17. RouteNote, Instagram carousels perform better than single photos according to the head of Instagram routenote.com
18. Social Media Today, Instagram Shares Algorithm Insights To Inform Strategy socialmediatoday.com
19. Hootsuite, Experiment, Carousels vs Reels blog.hootsuite.com
20. PostEverywhere, Instagram Carousel vs Reels posteverywhere.ai
21. TryMyPost, Instagram Carousel Algorithm 2026 trymypost.com

NEXT STEP

See where your brand stands.

Your Digital Authority Score, who AI recommends for your target questions, and the placements that close each gap. Your first report is free.

Get your free report →