Why you cannot see competitor Instagram saves

Saves are the most useful number Instagram will give you. A like costs nothing and means nearly nothing. A save is somebody deciding your post is worth returning to, and Instagram has said it uses saves and shares as ranking signals (Instagram’s ranking explainer, May 2023). How much weight is not published; the API reports the count, not what the ranking does with it.

Which makes it frustrating that you can see saves for exactly one account in the world: your own.

The rule, stated plainly

Saves on Instagram are the number of times accounts saved a post to return to it later. Meta reports saves only to the account that owns the post and has connected an app; no permission level exposes them for another account, so any competitor saves figure is modelled, bought or scraped.

Reach, views, saves, shares and reposts are private metrics. Meta releases them only to the account that owns the media, and only after that account has connected and authorised an application.

There is no permission, no review tier, no partnership and no price that exposes them for an account you do not manage; what any tool can measure stops at the public surface. This is not a gap in the API that a better-funded tool has solved. It is the design.

As of September 2026.

What changed in April 2026, and what did not

On 22 April 2026, Meta added three fields to the Instagram media endpoint for apps using Facebook Login: reposts_count, saved_count and shares_count. They apply across all API versions and are requested as ordinary fields rather than through the insights edge (Instagram Platform changelog).

This was a genuine improvement (what the April 2026 change actually added), and it was widely read as Meta opening up distribution data (Social Media Today). reposts_count in particular was not previously available in any form.

What it did not do is change who can see them. They remain private metrics available only for connected accounts, which is exactly how tools that integrated them describe the limitation (Socialinsider). The change made saves easier to collect for accounts that had already authorised you. It did not make a competitor’s saves visible.

That distinction gets lost in summaries, and it is the whole substance of the update. If you read in April that “saves are now available in the API” and concluded that competitive save data was coming, it was not, and, as of September 2026, it has not.

Why this quietly breaks competitive analysis

Here is the part that matters for anyone producing a monthly report.

For your account, connected, you can compute engagement rate over reach: interactions divided by the number of accounts that actually saw the post, with saves and shares included in the interactions.

For a competitor, you have likes and comments over follower count. Nothing else is obtainable.

These produce wildly different numbers from identical underlying performance. A post that reached 20% of an account’s followers and was saved heavily will score one way on the first measure and another entirely on the second. Put your reach-denominated 6.1% next to a competitor’s follower-denominated 1.3% and you have not discovered that you are outperforming them. You have discovered that you divided by different things.

The failure mode is not subtle once you look for it, and it is extremely common in competitive benchmark screens, because the interface has one column labelled “engagement rate” and two very different pipelines feeding it.

The only honest competitive comparison is like-for-like: your public metrics against their public metrics, follower-denominated on both sides. That number is less flattering and considerably more useful.

How to read a tool that shows you competitor saves

Some tools do display save or reach figures for accounts you do not manage. They are not necessarily lying, but they are doing one of three things, and you are entitled to know which:

1. Modelling. Inferring saves from a relationship between likes, comments and typical account behaviour. This can be directionally useful and is not a measurement.
2. Buying. Licensing data from a panel or a provider that obtained it some other way, with the coverage and staleness that implies.
3. Scraping. Collecting from surfaces not intended for it. This breaks, gets accounts restricted, and cannot survive Meta’s app review.

The question to put to any vendor is short: is this figure measured or derived? A tool that can answer that per metric is being straight with you. A tool that cannot, or that treats the question as impertinent, is telling you something anyway.

What to do instead

You cannot get competitor saves. You can get further than most people assume without them.

Compare your own account against itself. Your reach-denominated engagement rate this month against last month is the most informative number available to you, and it needs nobody’s permission.

Use comments as the public proxy for depth. Comment-to-like ratio is visible for every public account and correlates with the kind of investment that saves also measure. It is imperfect. It is real, which the alternative is not.

Watch posting patterns rather than performance. Format mix, cadence, caption length and hook structure are all fully public. What a competitor is doing is observable even when what they earned is not, and it is frequently the more actionable half.

State the denominator on every figure you publish. If a report contains one reach-denominated rate and one follower-denominated rate, they need different column headings, not the same one. Somebody will eventually compare them, and that is the point at which an honest report turns into a wrong decision.

The short version

Saves and shares are private. April 2026 made them easier to collect for accounts that connect, and changed nothing about accounts that do not. Any competitor save figure you are shown is modelled, bought or scraped.

The number is not available. A tool that shows a gap where the number should be is being accurate. A tool that fills it in without saying so has made your report worse in a way you cannot see.