Skip to content
All posts

How to read Instagram Insights like a coach

By Waris Hussain
Last updated 15 November 2026 · 5 min read

To read Instagram Insights like a coach, turn totals into rates: sends per reach, likes per reach, saves per reach and follows per reel. Split reach into followers and non-followers, read skip rate with average watch time, and compare every post to your own median. Review weekly and change one thing at a time.

Insights gives you dozens of numbers and no opinion. A coach's job is the opinion: which numbers matter, what they mean together, and what to do next. I build Cleo to do that job, and you can also do it by hand.

Which numbers matter most?

Start with the three Instagram itself points to. In January 2025, Adam Mosseri said the top three ranking signals are watch time, likes and sends, and told creators to watch "average watch time, likes per reach, and sends per reach" in their Insights (Social Media Today, 22 Jan 2025). In an October 2026 interview clip, he put sends per reach first in the list for new creators, then likes per reach, then watch time (kabo.sh summary, linking the clip).

Add a few more that tell you about growth and hooks, and the core set is:

Metric Formula What it tells you
Sends per reach Shares ÷ reach Do strangers value it enough to send it on?
Likes per reach Likes ÷ reach Did viewers enjoy it?
Average watch time From Insights Did it hold attention?
Skip rate From Insights Did the hook work?
Saves per reach Saves ÷ reach Is it worth coming back to?
Follows per reel Follows ÷ reach (or per 1,000 reach) Did it grow the account?
Non-follower share Non-follower views ÷ views Is it reaching new people?

Where do you find these in the app?

Instagram moves and renames Insights labels often, so match by meaning. Here's what Instagram has confirmed:

  • Skip rate is the share of views where people scrolled past in the first 3 seconds. It replaced "view rate" in August 2025, alongside a retention chart showing where viewers drop off (Social Media Today, 24 Aug 2025).
  • Followers vs non-followers: since January 2025, reel views can be split between followers and non-followers (TechCrunch, 29 Jan 2025).
  • Interactions: reel Insights list follows, likes, comments, saves and shares. Sends show up as shares. Accounts reached counts unique accounts that saw the reel on screen, played or not (Instagram Help Center, checked 15 Nov 2026).

Most of the rates you calculate yourself from these counts. One vendor reports that an April 2026 update split Insights into Overview, Engagement and Audience tabs and added a share rate (Inrō, updated 23 Jul 2026); if your app shows it, use it.

Why turn totals into rates?

Totals grow with views. A post that reaches more people will almost always have more likes, more comments and more sends, even if people liked it less. Rates remove that size effect, so you can compare a small post to a big one fairly.

Which numbers should you read together?

Single metrics can mislead, so read them in pairs:

Pair Pattern Likely meaning
Skip rate + watch time Low skip, low watch time Hook over-promised
Reach + follows High reach, low follows Profile or topic mismatch (fix)
Non-follower share + sends Low share, low sends Not travelling beyond followers
Saves + sends High saves, low sends Useful to save, less worth sending

How does a coach compare two reels?

Put two of your own reels side by side and fill in the rates:

Your stronger reel Your weaker reel
Reach your number your number
Shares (sends) per reach shares ÷ reach shares ÷ reach
Likes per reach likes ÷ reach likes ÷ reach
Saves per reach saves ÷ reach saves ÷ reach
Follows per 1,000 reach follows ÷ reach × 1,000 follows ÷ reach × 1,000
Non-follower share from Insights from Insights
Skip rate from Insights from Insights

Read this table by its rates, since a bigger reach on its own says little. If one reel was sent far more often per person reached and most of its views came from non-followers, it travelled. If the other lost more viewers in the first 3 seconds and barely left your followers, its hook is the likely problem. If saves per reach are close, the weaker reel's topic was still useful, so the fix is the hook and how shareable it is. Next week's one change: reuse the weaker reel's topic with a hook shaped like the stronger one's.

A 15-minute weekly review

  1. Pull last week's posts into a simple sheet.
  2. Calculate the rates above.
  3. Compare each to your median for the last 20 posts.
  4. Name the best and worst post, and one reason for each.
  5. Pick one change for next week, and only one.

Can you compare your rates to published benchmarks?

Rarely, and with care. Benchmarks often measure something different from what you're tracking. Socialinsider's 2026 Instagram benchmarks, for example, cover 67.5 million posts from business accounts with 1K to 1M followers, and calculate engagement as likes plus comments divided by followers (Socialinsider, 20 Feb 2026). That's a per-follower rate. The likes per reach and sends per reach Mosseri points to are per-reach rates, so the two can't be lined up directly.

What should you ignore?

  • Single-day spikes. Judge on weeks.
  • Other creators' engagement rates. Their audiences and norms differ, and they often use a different formula.
  • Comment counts on their own. Comments aren't among the three signals Mosseri named.
  • Benchmarks from blog roundups, unless you can trace them to a source.

This weekly review is what Cleo, Creator Kitchen's content coach, does for you automatically: she reads your Insights, calculates the rates, compares them to your own history, and tells you the one thing to change next, with the evidence on screen. Join the waitlist.

Sources