Why carousels get more saves than single posts

What does the data actually say about carousels versus single posts?

Two independent 2026 studies, both large enough to take seriously, agree that carousels beat single images and edge out Reels on engagement rate.

Socialinsider analysed 35 million posts across 447,613 active profiles and reported engagement rates of 0.55% for carousels, 0.52% for Reels and 0.37% for single images. Worth noting honestly: that report is published as a 2026 benchmark but analyses 2025 data, a labelling choice Socialinsider itself flags in the report.

Metricool’s 2026 study looked at 24,364,803 posts across 375,118 accounts and found carousels earn about nine times the saves of single images, roughly double the reach and views, and around four times the interactions. Over the same period single-image posts declined sharply year on year — reach down about 22%, interactions down about 25%.

Why does the carousel format itself drive saves?

Because a save is what people do with information they cannot finish consuming right now, and a carousel is the only feed format that reliably contains more than one thing.

A single image is either understood or scrolled past — there is nothing left over to come back to. A ten-slide breakdown of a process, a checklist, or a set of examples has residual value after the first read, and residual value is precisely what the save button captures.

There is a mechanical contribution too. Swiping is an interaction, so a carousel accumulates engagement from people who never tap like. And time on post scales with slides read, which gives Instagram more evidence that the post held attention.

Are saves really Instagram’s most important ranking signal?

No — and this is the most widely repeated falsehood in Instagram marketing content. In January 2025 Adam Mosseri named the top three ranking signals explicitly: watch time, likes, and sends. Saves were not among them.

His words: "The top three signals that matter most for ranking are watch time, likes and sends. So when looking at your insights, pay close attention to average watch time, likes per reach, and sends per reach." He added that likes matter slightly more for connected content and sends slightly more for unconnected content.

Saves do appear on Instagram’s official ranking explainer as one of several activity signals, so they are not irrelevant. But if you are optimising for one thing, the officially stated answer is sends — the DM share — not saves. Any guide telling you saves are the number one signal is repeating a claim Instagram’s own head has contradicted.

Also treat "DM shares outweigh likes by 3–5×" as fabricated. Mosseri said "slightly more important". No multiplier was ever attached to it by anyone at Instagram.

Which carousels get saved, and which do not?

Saved carousels have a reason to be returned to. That is the whole test, and it separates the format from the outcome — posting a carousel does not earn a save, posting something worth re-reading does.

• Reference material: specs, checklists, dimensions, shortcuts, templates. People save what they will need again and cannot memorise.
• Process breakdowns: a sequence of steps someone intends to follow later, which is a save precisely because they are not doing it right now.
• Curated collections: tools, examples, resources — value that is in the assembly rather than in any one item.
• Things worth showing someone else. This overlaps with sends, which is the signal Instagram actually weights most heavily for unconnected reach.
• What does not get saved: opinions, announcements, and motivational statements. They can perform well on likes and comments and almost never earn a save, because there is nothing to come back for.

Does this mean you should stop posting single images?

Not entirely, but the data makes it hard to justify them as a default. Single-image performance declined materially year on year in Metricool’s data while carousel and Reel performance held up, and a single image costs almost as much effort as a three-slide carousel.

The honest exception is anything where the image is the whole point — a photograph, a product shot, a piece of design. Splitting a genuinely singular image into slides to game the format is worse than posting it as it is.

The better framing: when you have more than one thing to say, the carousel is now the obviously correct container, and it was not always. When you have exactly one thing to show, show it.

How do you write a carousel people save rather than just read?

Front-load the promise and back-load the reason to return. A carousel that people finish and immediately forget usually gave everything away in the first two slides; a carousel people save usually contains something denser than they can absorb at scroll speed.

Concretely, that means putting the reference material — the numbers, the checklist, the specs — into slides rather than the caption, because saved posts are re-opened at the slides. And it means asking. A final slide that says "save this for the next time you build a deck" converts a meaningful share of readers who would otherwise have swiped away satisfied.

One caution about statistics: the widely circulated "carousels get 1.4× more reach and 3.1× more engagement" figure comes from a 2020–2022 era study and is contradicted by the current numbers above. If you quote carousel performance, quote a study with its sample size and year attached — a habit that also happens to be why an AI assistant would cite you rather than the fiftieth blog repeating an eight-year-old stat.