Social media engagement metrics explained

What engagement rate, reach, views, saves, shares, watch time and click-through rate measure, with formulas, worked examples and why benchmarks rarely transfer.

Key points

  • Engagement rate has at least three common definitions (by followers, by reach and by views or impressions), and each gives a different number for the same post.
  • Reach counts unique accounts; views or impressions count every display, so the ratio between them shows how often people saw the content.
  • Saves, shares and watch-time metrics usually say more about content quality than likes do.
  • Benchmarks vary because of definitions, account size, format, industry and platform changes, so compare against your own history first.
  • Decision metrics are the ones that change what you do next; vanity metrics look good in a report but do not.

“Engagement” is the most quoted and least consistently defined number in social media marketing. Two people can look at the same post and report engagement rates of 4% and 10%, and both can be correct, because they are dividing by different things. This explainer sets out what the main metrics measure, how to calculate them, and which ones should influence decisions for a business selling online.

The building blocks: reach, views and impressions

Before any rate can be calculated, you need to know what the denominator is.

Reach is the number of unique accounts that saw a piece of content at least once. Meta’s Instagram metrics reference defines it as the number of unique Instagram users who have seen the media at least once, and marks it as an estimate.

Impressions traditionally meant the total number of times content was displayed, including repeat displays to the same person. Views is the term Meta and TikTok now use for a similar idea. On Instagram, Meta’s reference describes views as the total number of times the media has been played and lists impressions as deprecated for media created after 2 July 2024. Facebook made a similar switch to views as its headline metric. The practical effect is that historical reports using “impressions” may not be directly comparable with current reports using “views”.

The relationship between them is frequency:

Frequency = views (or impressions) ÷ reach

Illustrative example: a Reel with 5,000 views and a reach of 4,000 has a frequency of 1.25, meaning the average viewer saw it one and a quarter times. Organic content usually has a frequency close to 1; paid campaigns can push it much higher, which is worth watching because repeated exposure eventually stops working.

Engagement: what counts

An engagement (or interaction) is an action someone takes on the content. Meta’s reference defines total interactions as likes, saves, comments and shares, net of unlikes, unsaves and deleted comments. Other tools add link clicks, profile visits, replies to Stories or follows. Before comparing figures from different sources, check which actions they include.

For the worked examples below, assume a single Instagram Reel from an account with 10,000 followers:

Metric Value
Followers 10,000
Reach 4,000
Views 5,000
Likes 300
Comments 40
Saves 30
Shares 30
Total engagements 400

Illustrative figures chosen for easy arithmetic.

Three ways to calculate engagement rate

Engagement rate by followers

ER (followers) = engagements ÷ followers × 100

400 ÷ 10,000 × 100 = 4%

This is the most widely quoted version because followers are public, so third-party tools can calculate it for any account. Its weakness is that it assumes every follower had the chance to see the post, which is rarely true, and it ignores non-followers who saw the content through recommendations.

Engagement rate by reach

ER (reach) = engagements ÷ reach × 100

400 ÷ 4,000 × 100 = 10%

This measures how compelling the content was to the people who actually saw it. It is usually the most useful version for judging content, but it needs access to the account’s private insights.

Engagement rate by views or impressions

ER (views) = engagements ÷ views × 100

400 ÷ 5,000 × 100 = 8%

This version is common in paid-media reporting. It falls as frequency rises, because repeat displays add to the denominator without necessarily adding engagements.

The same post has produced 4%, 10% and 8%. None is wrong. The important thing is to state which formula you used and to use the same one every time. If you want to check figures quickly, HowSociable’s engagement rate calculator runs these calculations from the inputs you provide.

Averaging across posts

When reporting engagement for a month, there are two methods. You can sum all engagements and divide by total reach (a weighted average), or calculate each post’s rate and take the mean (a simple average). They give different results when posts vary widely in reach. The weighted method better reflects overall performance; the simple average stops one viral post dominating. Again, pick one and stick with it.

Saves and shares

Not all engagements are equal. A like costs the user almost nothing. A save suggests someone intends to come back to the content, which is a useful signal for products with a longer consideration period, such as furniture, recipes or outfits. A share puts the content in front of new people and implies a personal recommendation.

Calculating these separately is often more informative than a combined rate:

Save rate = saves ÷ reach × 100

In our example: 30 ÷ 4,000 × 100 = 0.75%. The share rate is the same here, also 0.75%.

Track these over time for your own account. A product video with a low like count but a rising save rate may be doing more commercial work than a meme with many likes.

Watch time and completion

For video, attention matters more than reactions. TikTok’s explanation of how the For You feed works names watching a longer video to the end as a strong signal of interest, and its analytics tutorial lists average watch time among the per-video measures. Instagram exposes an equivalent average watch time for Reels.

Average watch time = total play time ÷ views

Average percentage viewed = average watch time ÷ video length × 100

Completion rate = views that reached the end ÷ total views × 100

Illustrative example: a 30-second product demo with 5,000 views and 50,000 seconds of total play time has an average watch time of 10 seconds, or 33% of its length. If 750 views reached the end, the completion rate is 15%.

Platforms define these slightly differently (some count a view as “complete” if it reaches the last few seconds, and replays can inflate total play time), so use them for comparing your own videos rather than for comparing against other accounts. A steep drop in the first two or three seconds usually points to the opening, not the product.

Click-through rate

Click-through rate (CTR) measures how often seeing something led to a click.

CTR = clicks ÷ impressions (or views) × 100

If the Reel above carried a product link that received 80 clicks: 80 ÷ 5,000 × 100 = 1.6%.

Definitions vary by platform. YouTube’s impressions and click-through-rate FAQ, for instance, defines impressions click-through rate as how often viewers watched a video after seeing a registered impression on YouTube, not clicks to an external website. In ad platforms, check whether a figure is “link CTR” or “all clicks”, which includes clicks on the profile name, the “more” button or the image.

The next step happens on your site. Clicks recorded by the platform will usually exceed sessions recorded by your web analytics, because of in-app browsers, slow page loads and visitors who decline analytics cookies. Our guide to using analytics to improve ecommerce conversion covers what happens after the click.

A note on “engagement rate” in Google Analytics

Google Analytics 4 uses the same phrase for something unrelated. In GA4, engagement rate is the percentage of website or app sessions that were “engaged”: lasting longer than 10 seconds, including a key event, or including two or more page or screen views. A social engagement rate and a GA4 engagement rate cannot be compared or combined.

Why benchmarks vary so much

Published “average engagement rates” for each platform differ widely, and we do not quote them here. The reasons are structural:

  • Different formulas. A benchmark by followers will look very different from one by reach.
  • Account size. Smaller accounts tend to show higher engagement rates by followers, because a larger share of a small audience is actively interested.
  • Format. Carousels, Reels, Stories and static posts behave differently, and some benchmarks mix them.
  • Industry and audience. A local bakery and a software company should not expect the same response.
  • Data access. Third-party benchmark studies usually rely on public data, which means follower-based rates and no saves or reach.
  • Platform changes. Ranking systems change continually; Meta’s ranking explainer describes how feeds combine followed and recommended content using AI predictions. Metric definitions also change, as the move from impressions to views showed.

In our view, the most reliable benchmark is your own account’s median over the previous three to six months, broken down by format. A post that beats your own median is doing well, whatever an industry average says.

Vanity metrics versus decision metrics

A useful test for any metric is whether a change in it would change what you do.

Vanity metrics are numbers that look good in a report but do not lead to a decision: total follower count, total likes, raw views on a post that was not designed to sell anything. They are not useless, but they are context.

Decision metrics connect to an action:

  • Average percentage viewed tells you whether to re-cut the opening of a video.
  • Save and share rates tell you which topics to make more of.
  • CTR tells you whether the call to action or link placement is working.
  • Sessions, add-to-baskets and orders from social traffic, measured in your own analytics, tell you whether the channel earns its keep.
  • Engagement by reach on creator content helps decide which partnerships to renew; see our guide to measuring influencer campaign effectiveness.

Which of these matter most also depends on the platform’s discovery model. On interest-graph feeds, watch time and completion carry more weight; on follow-graph surfaces, reach among existing followers matters more. Our comparison of Instagram, TikTok and Facebook for product discovery explains the difference.

Bottom line

Choose one definition for each metric, write it down, and use it consistently. Prefer reach-based engagement for judging content, watch time and completion for video, and on-site sessions and orders for judging commercial results. If your current reporting tools make this difficult, our comparison of ecommerce analytics tools may help. Compare against your own history before anyone else’s, and give the most weight to the numbers that tell you what to do next.

Sources and further reading

  1. Instagram Media Insights (metrics reference) — Meta for Developers
  2. TikTok Tutorial Series: TikTok Analytics — TikTok Newsroom
  3. How TikTok recommends videos #ForYou — TikTok Newsroom
  4. Impressions & click-through-rate FAQs — YouTube Help
  5. [GA4] Engagement rate and bounce rate — Google Analytics Help
  6. Our approach to explaining ranking — Meta Transparency Center