Instagram reporting needs to separate public engagement from private movement

The most common Instagram report still compresses too many behaviours into one engagement rate. That worked better when the feed was simpler and public reactions carried more of the visible signal. In 2026, the useful movement is split across saves, shares, comments, reposts, profile visits, messages and later conversion paths. A single engagement percentage can hide the actual job a post performed.

Socialinsider's 2026 Instagram benchmark reports weaker 2025 engagement across its global brand-post sample and shows format differences between Carousel, Reels and images. Its separate Instagram engagement report, based on 15 million posts, breaks the problem open: Reels have stronger median comments and shares, while Carousel has stronger saves. Metricool's Instagram Study 2026 adds a larger connected-account view, but also carries the important caveat that Instagram's metric definition shifted from impressions to views in April 2025.

Instagram's own product direction reinforces the need to separate behaviours. In its August 2025 announcement on Reposts, Friends tab and Map, Instagram says reposts preserve original creator credit and may be recommended to the reposter's followers. Friends tab and Map are not simple public-like mechanisms; they point to distribution through social context, location and relationship signals. Availability varies by market and feature, so the announcement should be treated as a product signal rather than a universal reach guarantee.

The monthly report should show movement type

For planning, split Instagram reporting into four lanes. The first is public response: comments and visible reactions. The second is private or semi-private movement: shares, sends, reposts and message starts. The third is retrieval: saves, repeat visits and content that reappears in sales or customer-service conversations. The fourth is conversion support: profile action, catalogue movement, link clicks, lead form starts, message quality and revenue where attribution is available.

This is not more reporting for its own sake. It prevents the wrong optimisation. A post with modest likes but high saves may be a strong explainer. A Reel with low saves but high shares may be doing awareness work. A product post with weak public engagement but strong message starts may be a commercial asset. One engagement rate makes these three posts look comparable when they are not.

What should change

The dashboard should not rank every post by one aggregate number. It should tag the intended job before publication, then review performance against that job. A Reels asset built to spark conversation should not be penalised because it was not saved. A Carousel built to support decision-making should not be penalised because it was not heavily commented on.

For reporting discipline, rebuild the monthly Instagram report around movement type: public response, private distribution, retrieval and conversion support. Keep engagement rate as a reference number, not the steering metric.