The Things You Should Know About Instagram’s Algorithm Change and Its Effect on Your Engagement Rate

First, there is not really one new Instagram algorithm

We tend to talk about “the algorithm” as if Instagram has one mysterious machine deciding whether your latest Reel deserves 20,000 views or 200.

It is more complicated than that.

Instagram uses different ranking systems across Feed, Reels, Stories, Explore and recommendations. Those systems are designed around how people behave in each part of the app. Current guidance summarised by Hootsuite makes the same point, different areas of Instagram rank content differently. (Social Media Dashboard)

What has become clearer is what Instagram increasingly values.

Instagram head Adam Mosseri has said the three signals creators should pay particular attention to are watch time, likes and sends. He has also explained that likes matter slightly more when Instagram is deciding what to show your existing followers, while sends matter slightly more when deciding what to recommend to people who do not follow you. (Social Media Today)

That distinction matters.

A piece of content can be perfectly popular with your existing community without necessarily travelling very far beyond it.

Engagement really has fallen

This is the bit worth acknowledging properly.

Socialinsider analysed 35 million Instagram posts and reported average engagement of around 0.48 percent in its 2026 benchmark data, representing a roughly 24 percent year-on-year decrease. Its quarterly data also shows Instagram engagement averaging around 0.45 percent in the second quarter of 2026. (Socialinsider)

Buffer saw a similar direction of travel.

Its 2026 engagement report found Instagram’s median engagement rate fell from roughly 7.3 percent in 2024 to 5.4 percent in 2025, around a 26 percent decrease. Buffer calculates engagement against reach, which is why its absolute percentages look very different from Socialinsider’s follower-based numbers. (Buffer)

That is an important detail.

Different studies use different formulas, account samples and definitions of engagement. You should not put their raw percentages next to each other and assume they are measuring exactly the same thing.

What you can compare is the direction.

And the direction is fairly clear.

Engagement has become harder to earn.

Reach has been feeling the pressure too

Engagement is only part of the picture.

Metricool’s 2026 Social Media Study analysed almost 40 million posts and found that Instagram post reach fell by 31 percent, while Reels reach dropped by 35 percent year on year. (Metricool)

That might sound dramatic, but the context matters.

There is simply much more content competing for attention.

Brands are publishing more. Creators are publishing more. AI tools have made content production faster. Short-form video is no longer a novelty. Every scroll now contains an enormous amount of competition for the same few seconds of attention.

Metricool’s later Instagram-specific study adds another useful layer. It found that brands were publishing around 24 percent more content, while total views and interactions were actually increasing. Overall engagement rate, however, slipped from 3.79 percent to 3.54 percent. (Metricool)

In other words, people have not disappeared.

Their attention has simply become more divided.

Not every type of content is suffering equally

This is where the conversation gets more useful.

If Instagram engagement as a whole is declining, it does not automatically follow that every format is performing badly.

Socialinsider found carousels remained its strongest Instagram format, averaging around 0.55 percent engagement in its annual benchmark data. Reels followed closely, while single images performed considerably lower. (Socialinsider)

Buffer found an even bigger gap when measuring engagement against reach. In its dataset, carousels generated a median engagement rate of 6.9 percent, compared with 4.44 percent for single images and 3.31 percent for Reels. Buffer also points out that Reels frequently have a different job, generating views and discovery rather than necessarily producing the highest interaction rate. (Buffer)

Metricool’s Instagram study paints a particularly difficult picture for single-image posts. It found their reach fell by nearly 22 percent year on year, interactions dropped by roughly 25 percent and engagement fell by almost 46 percent. At the same time, carousels generated substantially more saves, while Reels continued to produce far more interactions than single-image posts. (Metricool)

So rather than saying “Instagram engagement is dead”, a more useful summary would be this.

• Passive, single-image content is having a tougher time

• Carousels are proving particularly resilient

• Reels still matter for discovery, but simply posting a Reel no longer guarantees reach

• Shares, saves, watch time and meaningful interactions are becoming more useful indicators than likes alone

Instagram is becoming more interested in what people actually do with your content

There has also been a noticeable shift away from treating a like as the ultimate sign of success.

Think about the difference between these two actions.

Someone sees your post, taps the heart and continues scrolling.

Someone sees your post and immediately sends it to a colleague because it is useful, funny, painfully accurate or worth discussing.

The second behaviour tells Instagram much more.

That helps explain the growing attention around sends, watch time and content people actively want to pass on.

Socialinsider has also found some interesting changes underneath the headline engagement number. Its broader 2026 benchmark research reports a 16 percent fall in average Instagram comments, while shares per post increased by 12 percent and video views increased by 29 percent year on year. (Socialinsider)

People may be interacting differently, not simply interacting less.

And that means an old engagement rate formula may not tell you everything you need to know anymore.

Original content matters more too

Another significant change arrived in April 2026.

Instagram expanded its protections for original creators, restricting recommendation reach for accounts that primarily repost photos and carousels created by other people. Similar protections already existed around Reels. The change affects recommendations to new audiences rather than simply preventing followers from seeing an account they already follow. (TechCrunch)

For most brands creating their own photography, graphics, videos, commentary and campaigns, this should not be a reason to panic.

If anything, it creates another reason to prioritise content with a recognisable point of view rather than filling a feed with lightly repackaged material from elsewhere.

So what should you actually do

Probably not reinvent your entire Instagram strategy overnight.

Instead, this is a useful moment to update what you measure and what you expect from individual posts.

Look beyond likes and consider things such as

• Shares and sends

• Saves

• Average watch time

• Reach from non-followers

• Carousel performance

• Views

• Profile visits and actions

• Engagement relative to your own recent baseline

It is also worth comparing like with like.

Do not judge this month’s performance against a particularly viral month two years ago and conclude something has gone catastrophically wrong.

Compare recent Reels with recent Reels. Compare carousels with carousels. Look at the last three or six months. Look for formats, topics and creative approaches that consistently sit above your new normal.

That is much more useful than chasing a universal “good engagement rate”.