Instagram Reels Insights: How to Read Every Metric

Reels Insights is Instagram's free, per-video analytics panel. It answers two questions: how far the Reel travelled, through views, reach and where those views came from, and how people responded once it reached them, through watch time, engagement and what they did next. Most people check the view count and stop there, which is the one number that tells you least.

How to open Reels Insights

1. Go to your profile and open the Reels tab.
2. Tap the Reel you want to look at.
3. Tap View insights under the Reel.

For several Reels at once, use the Professional Dashboard instead. From your profile, open Professional Dashboard, then Content, and filter to Reels. That view lets you sort by metric and compare Reels side by side, which is where patterns show up that a single Reel never reveals.

If you cannot see insights at all, the usual cause is account type. Full per-Reel insights require a professional account, Creator or Business, rather than a personal one. Switching is free and takes a minute in Settings. Very old Reels also predate the analytics, so nothing will appear for those.

Views and accounts reached are not the same number

Views counts plays, including repeat plays. One person watching three times is three views.

Accounts reached counts unique accounts that had your Reel on screen, whether or not they watched it. Someone scrolling past still counts.

The gap between the two is the useful part. If reach is 10,000 and views are 18,000, people are rewatching, which is a strong signal on short or dense content. If views barely exceed reach, most people saw it once and moved on.

Watch time and the retention curve

Watch time is the total time played across everyone. Average watch time divides that by first-time plays, so it approximates how long a typical new viewer stayed.

The number is meaningless without the Reel's length beside it. Twelve seconds average on a fifteen-second Reel is excellent. Fifteen seconds average on a sixty-second Reel means the back half is not working. Always read it as a proportion, and never against another creator's number.

The retention chart is the more actionable of the two, because it shows the percentage still watching at each second rather than one average. A normal curve falls sharply in the first second or two as people scroll by, then flattens once the remaining viewers have committed, then tails off.

What you are hunting for is a sharp drop somewhere in the middle. That is a specific moment losing people: a slow edit, a transition that confuses, or the point where the video stops delivering what the opening promised. An average cannot tell you that. The curve can point at the second.

The first three seconds

Instagram has been consistent in its guidance to creators that the opening seconds matter more than anything else, and its own insights reflect that: the metrics it surfaces distinguish between people who scrolled straight past and people who kept watching beyond the first few seconds.

Treat that number as a hook diagnostic rather than a verdict on the whole Reel. A weak opening will drag it down even when the content that follows is good, and the fix is almost always the first line, the first frame or the first bit of on-screen text rather than anything later.

If you want to test two openings against each other before committing a Reel to your followers, that is exactly what Reel trials are for: post both to non-followers only, compare the curves, then publish the one that held people.

One caution worth stating: you will find articles quoting exact thresholds for what counts as a healthy figure here. Instagram does not publish them, and they vary enormously by niche, length and audience. Compare a Reel against your own recent Reels, which is a comparison that actually means something, rather than against a number someone invented.

Where your views came from

The source breakdown splits views across the surfaces they arrived on: your followers' feed, the Reels tab, Explore, search, your profile, and shares into Stories.

This is the clearest signal of whether you are growing or just being seen by the same people. Views dominated by the feed mean your existing followers liked it and it did not travel. Views dominated by the Reels tab mean Instagram pushed it to people who do not follow you, which is where new audience comes from. Search views mean the Reel is being found deliberately, which tends to keep earning long after the initial spike.

If the Reels tab is barely showing up in your sources, the issue is usually distribution rather than content, and it is worth understanding how Instagram decides what to show people before rewriting anything.

Engagement, and when it happened

Instagram now shows when during playback people tapped like, not just how many did.

Read it against your retention curve rather than on its own. Likes early usually come from people who already know you and did not need convincing. Likes clustered around a specific moment tell you which part actually landed, and that is the part worth building more content around. Likes only at the very end suggest the Reel earned its payoff but had nothing memorable before it.

On the other engagement types, Adam Mosseri has publicly pointed creators toward views and sends per reach as the two worth watching. Sends per reach is the useful one here, because sending a Reel to someone in a DM is a much higher bar than tapping a heart. Before you post, it is worth asking whether anyone would actually forward this to a friend. If not, the Reel will get likes and go nowhere.

Saves sit somewhere in between: they signal reference value, though plenty of people save things they never return to.

What people did afterwards

Follows, profile visits and link taps attributable to that specific Reel.

Follows per Reel is the honest measure of whether content is growing the account. A Reel with enormous views and almost no follows entertained people who do not want more from you, which is fine occasionally and a problem as a pattern. A Reel with modest views and a strong follow rate found the right audience, and is worth repeating.

Link taps will be low for most accounts, and that is usually a structural problem rather than a content one, since Reels cannot carry a clickable link unless you pay for Meta Verified or boost the post.

Who your Reel reached

The audience section splits reach between followers and non-followers, and gives age, gender and country for the accounts reached.

Neither split is inherently better. Mostly non-followers means the Reel worked as discovery. Mostly followers means it resonated with the people you already have. What matters is whether the split matches what you wanted that Reel to do.

The demographic data is worth checking when a Reel overperforms. If it reached a noticeably different age or country profile than your usual audience, you have found a new segment, and the decision is whether to pursue it or treat it as a one-off.

What to do with a Reel that performs

Insights tell you which Reel worked. The gap is what happens next, because the engagement a good Reel generates has a short shelf life.

The typical version: a Reel takes off, a few hundred people comment, you reply to a dozen over the following days, and the rest have long since moved on. The interest was real and it expired.

The alternative is to decide in advance what a commenter receives. Put a keyword in the caption, and anyone who comments it gets a DM within seconds containing whatever you promised, plus a contact record you can use later. Across Inrō's own automations, DMs average a 90% open rate, a 60% reply rate and a 28% click-through rate, which is the reason this is worth setting up before a Reel performs rather than after. The full setup is in generating leads from Instagram Reels, and the mechanics of what you can send are covered in sending links in Instagram DMs.

Insights tell you where to point it. A Reel with high non-follower reach is reaching strangers, so that is the one to attach a keyword to. A like spike at a particular moment tells you where in the video to put the call to action, because that is when people are paying most attention.

The same logic applies to Stories, where replies arrive as DMs rather than public comments. That distinction, and what it means for follow-up, is covered in Story replies versus Story comments.

Catch the engagement while it is still warm

Every commenter gets an instant DM and becomes a contact you can message again. Free plan, no card required.

Start for free

FAQs

Why can I not see insights on my Reels?

Almost always because the account is personal rather than professional. Full per-Reel insights need a Creator or Business account, which is free to switch to in Settings. Reels posted long before the analytics existed will also show nothing.

Can I see who watched my Reel?

No. Instagram gives you aggregate demographics and the follower versus non-follower split, but never individual identities. No account type changes this, and any tool claiming otherwise is not doing what it says.

What is the difference between views and accounts reached?

Views counts plays including repeats, so one person can generate several. Accounts reached counts unique accounts that had it on screen, even if they never pressed play. Views well above reach means people are rewatching.

What is a good average watch time?

There is no universal number, because it depends entirely on how long the Reel is. Read it as a proportion of total length, and compare against your own recent Reels rather than anyone else's.

How do I read the retention chart?

It shows the percentage still watching at each second. A steep initial drop is normal scroll-by. What you are looking for is a sharp fall partway through, which points at the exact moment that lost people, so you know what to change rather than guessing.

Are Reels Insights free?

Yes, entirely. It is built into Instagram, needs no third-party tool or subscription, and works on any professional account.

Is it safe to automate DMs from Reel comments?

Yes, when the tool connects through Meta's official API and only replies to people who acted first. A comment is that action. Anything that logs in as you rather than using the API is the risk to avoid, and the rules around what is allowed are worth reading before you set anything up.