# Instagram Reels metrics that matter: views, shares, saves and watch time

URL: https://ai-xodim.uz/en/blog/reels-analytics-metrics
Published: 2026-10-12 · Updated: 2026-10-23

Views, reach, shares, saves, watch time and engagement rate by reach: what each Reels metric means, how to compare reels fairly and what to do next.

- The Reels metrics that show whether a reel worked are shares, saves and watch time measured against reach; views only show how far a reel travelled.
- Engagement rate by reach — (likes + comments + saves + shares) ÷ reach — shows how the people who actually saw a reel reacted, while engagement rate by followers divides by people who may never have seen it.
- Compare each reel with the median of your own recent reels instead of “industry benchmarks”: twice the median is a winner, half of it is a miss.
- Once a month, sort your reels by shares per reach, compare the top three with the bottom three, write down what the leaders have in common and plan two experiments.

You posted twelve reels this month. One took off, three did fine, the rest quietly disappeared. The obvious question is “what should I make more of?” — and the number Instagram shows first, views, is the worst one to answer it with.

This guide covers the Reels metrics that actually tell you what worked, how to compare reels fairly, and how to turn a month of numbers into two or three decisions for the next one.

## Views vs reach: why the biggest number misleads

**Views** count every time your reel was played, including repeat plays by the same person. **Reach** counts the unique accounts that saw it. A reel with 50,000 views and 20,000 reach was seen by 20,000 people, and some of them watched it more than once.

Views answer one question: how far did the reel travel? They say nothing about whether people liked it, learned something or wanted to pass it on. Two traps to avoid:

- **Don't add up daily reach.** Reach is unique per period: someone who saw your reels on Monday and on Tuesday is one person for the week, not two. Use the 7-day or 30-day figure as Instagram reports it.
- **Don't compare raw numbers between reels with different reach.** 300 shares on a reel seen by 100,000 people is a weaker result than 150 shares on a reel seen by 10,000. Divide by reach first.

## The Reels metrics that actually matter

### Shares (sends)

A share means someone sent your reel to a friend in Direct or reposted it to their story. It's the clearest sign that the content is valuable to someone other than you — and Instagram has said publicly that how often a post is sent per reach weighs heavily in what it recommends. Track **shares ÷ reach**, not raw shares.

### Saves

A save means “I'll need this later”: a recipe, a checklist, a price list, a how-to. A high **saves ÷ reach** says the reel is useful, even if it isn't funny or emotional. Tutorials and lists usually win here.

### Watch time: average and total

**Average watch time** is how long a typical play lasted. Always compare it with the length of the reel: 9 seconds of a 10-second reel and 9 seconds of a 60-second one are very different stories. If most people leave in the first seconds, the problem is the opening, not the topic.

**Total watch time** is all the time people spent on the reel together. It shows how much attention the reel collected and is handy for comparing reels of different lengths. Instagram has also said that watch time is among the signals it relies on most for recommendations.

### Likes and comments

Still worth a look, but they are cheaper signals: a like costs one tap. Comments become valuable when they turn into conversations — for example, when a reel asks people to comment a word and a [comment-to-DM auto-reply](/en/blog/instagram-comment-to-dm) sends each of them the link.

## Engagement rate by reach vs by followers

Engagement rate (ER) puts interactions in proportion. There are two common versions:

- **ER by reach** = (likes + comments + saves + shares) ÷ reach × 100%
- **ER by followers** = the same interactions ÷ followers × 100%

Take a hypothetical reel with 1,000 interactions that reached 20,000 accounts, on an account with 5,000 followers. ER by reach is 5%; ER by followers is 20%. The second number looks impressive, but most of those 20,000 people weren't your followers — the formula divides by the wrong crowd. ER by followers inflates reels that travelled outside your audience and makes everything else look weaker than it is.

**ER by reach** answers the question that matters: of the people who actually saw the reel, how many reacted? That's why Ai Xodim's [Instagram analytics](/en/instagram-analytics) uses the **median ER by reach** across your posts as the main figure and shows ER by followers next to it only for comparison.

## Your own median beats “industry benchmarks”

You will find articles quoting one “good engagement rate for Reels”. There isn't one. A bakery in Tashkent, a fitness coach with 200,000 followers and a B2B consultant live in different worlds: niche, audience size, language and format all move the numbers.

The baseline that matters is **your own median** — the middle value of your recent reels when you line them up from lowest to highest. Unlike an average, the median isn't dragged up by one viral outlier. With it, every reel gets a simple label:

- **twice the median or more** — a winner: study it;
- **around the median** — your normal;
- **half the median or less** — a miss: find out why.

Ai Xodim's Reels report does exactly this: it compares each reel with the median of your recent reels on views, reach, shares, saves and watch time.

## Cheat sheet: metric → what it tells you → what to do

| Metric | What it tells you | What to do |
|---|---|---|
| Reach | How many unique people saw the reel | Compare hooks and topics among the reels that reached furthest |
| Views ÷ reach | Whether people watch it again | Clearly above 1? Consider a part 2 or a longer version |
| Shares ÷ reach | Whether it's worth sending to a friend | Repeat the topic and format of the leaders |
| Saves ÷ reach | Whether it's useful enough to keep | Turn it into a series, a checklist or a carousel |
| Average watch time | Whether the opening and pace hold people | People leave in the first seconds? Rewrite the first line |
| Total watch time | How much attention the reel collected | Use it to compare reels of different lengths |
| Median ER by reach | How people who see your posts react overall | Track it month to month, not reel to reel |

## How to run a monthly Reels review in 20 minutes

1. **Take the last 30 days of reels.** With fewer than six or so, wait a couple of weeks — a median of three reels says little.
2. **Sort them by shares ÷ reach.** Mark the top three and the bottom three.
3. **Check watch time** for both groups against each reel's length. Did the leaders hold people longer, or were they simply sent more?
4. **Write down what the leaders have in common:** topic, first sentence, format (talking head, tutorial, list, behind the scenes), length, time of posting.
5. **Name the reason for each miss:** a slow opening, a topic only you care about, no clear point, a weak ending.
6. **Pick two experiments** for the next month — for example, “open every reel with a question” and “two more tutorials on the top topic”.
7. **Make the next winner work for sales.** Add a keyword call to action (“comment GUIDE and I'll send it to you”) and set up a [comment auto-reply](/en/instagram-comment-auto-reply) that delivers it in Direct.

## 10 questions to ask your AI about your reels

Connect Claude, ChatGPT or Gemini to Instagram through the [Ai Xodim AI connector](/en/ai-connector), and your AI gets the Reels report in one request — views, reach, shares, saves, average and total watch time, rates by reach and the comparison with your median. Then simply ask:

1. “Which reels this month got the most shares per reach, and what do they have in common?”
2. “Which reels lost viewers in the first seconds? Compare their openings with the leaders.”
3. “Rank my last 15 reels by saves per reach and suggest three topics for a series.”
4. “Which reels were shared more than usual but got few likes? Why might that be?”
5. “Compare my tutorial reels with my behind-the-scenes reels on average watch time.”
6. “What is my median engagement rate by reach, and which posts are far above it?”
7. “Which reel brought the most comments, and what were people asking?”
8. “Which of my reels deserve a part 2?”
9. “Write three new opening lines in the style of my best-performing reel.”
10. “Set up an auto-reply for the word LINK under my latest reel.”

The last one doesn't just analyse — it acts. The AI shows you the exact rule first and saves it only after your “yes”. The Analytics tab in the dashboard is free; the AI connector is part of [Ai Xodim Pro](/en/pricing). If you haven't connected an assistant yet, here's [how to connect ChatGPT, Claude or Gemini to Instagram](/en/blog/connect-chatgpt-claude-gemini-to-instagram).

## Monthly checklist

- Use 7-day or 30-day reach — never daily reach added up.
- Sort reels by shares ÷ reach, then by saves ÷ reach.
- Mark reels at twice the median and below half of it.
- Compare each reel's average watch time with its length.
- Write down three things the leaders have in common.
- Name the reason for each miss.
- Plan two experiments for the next month.
- Put a keyword call to action under the next reel of the winning type.
- Note your median ER by reach to compare with next month.

## FAQ

### What is a good engagement rate for Instagram Reels?

There is no universal number: niche, audience size and format change it a lot. Compare each reel with your own median engagement rate by reach — twice the median is a strong reel, half of it a weak one.

### Why are my reel's views higher than its reach?

Views count every play, including repeat plays by the same person, while reach counts unique accounts. More views than reach means some people watched the reel more than once.

### Can I add up daily reach to get the monthly reach?

No. Reach counts unique accounts, so adding up days counts the same people several times. Use the 7-day or 30-day reach instead.

### Which Reels metric matters most?

Shares per reach and watch time tell you the most: Instagram has said both weigh heavily in recommendations. Saves per reach show how useful the content is.

### Is Ai Xodim's Instagram analytics free?

Yes, the Analytics tab in the dashboard is free. The Reels report through your own AI — Claude, ChatGPT or Gemini — is part of the AI connector in Ai Xodim Pro: $9.99 a month, the first 3 days free, no card needed.