Marketing
Social Media Metrics That Matter, Network by Network
Bluesky and Mastodon count no views, X keeps link clicks for 30 days, and LinkedIn calls its impressions an estimate. Here's what to track on each network.
Oleh Husiev8 min read
The social media metrics that matter are the ones every network reports and that track a goal: replies and reposts per post, follower growth rate, and link clicks you measure yourself with UTM tags. Engagement rate only works where a network counts views, and Bluesky and Mastodon don't count any.
Most metrics advice assumes every network gives you the same numbers. They don't. Here's what Threads, X, LinkedIn, Bluesky and Mastodon actually report, what that means for comparing them, and a short routine for reading the numbers you do get.
What each network's analytics actually report
This is what each network gives you, or any tool you connect, for your own posts. It comes from each network's own documentation, checked on November 16, 2026.
| Network | Views per post | Interactions per post | Clicks | Audience |
|---|---|---|---|---|
| Threads | Views (Meta marks this metric as still in development in its API) | Likes, replies, reposts, quotes | Link clicks, for the whole account | Followers, profile views; demographics from 100 followers |
| X | Impressions, public on every post; views on videos | Likes, replies, reposts, quotes, bookmarks | Link clicks and profile clicks, your own posts only | Followers |
| Impressions (LinkedIn calls the number an estimate) and members reached | Reactions, comments, reposts; saves and sends on personal posts | Clicks on Page posts | Share of impressions from inside and outside your network | |
| Bluesky | None | Likes, replies, reposts, quotes, bookmarks | None | Followers |
| Mastodon | None | Favourites, boosts, replies, quotes (from version 4.5) | None | Followers |
A few details change how you read these:
- Threads demographics (country, city, age and gender) only appear once an account has at least 100 followers, and insights through the API go back to April 13, 2024.
- X only gives tools link clicks, profile clicks and other private metrics for posts from the last 30 days. After that, a tool reading X can still see the public counts, but not the clicks.
- LinkedIn splits impressions into in-network (your connections and followers) and out-of-network. Members reached counts each person once, so it's lower than impressions. On a Page, metrics other than reactions and comments can take up to 48 hours to show.
- Bluesky has had an open request for view counts on its app repository since January 2025. Until that ships, there's nothing to divide by for an engagement rate.
- Mastodon doesn't count views at all, on any server.
Why engagement rate doesn't compare across networks
Engagement rate is interactions divided by views:
Engagement rate = (likes + replies + reposts) / views x 100
It's a good number for one network over time. It falls apart across networks for three reasons:
- Two networks have no views. On Bluesky and Mastodon there's nothing to divide by, so any "engagement rate" a tool shows there is something else, usually a score built from interactions alone.
- A view means different things. Threads counts every time a post is played or displayed. LinkedIn calls its impression count an estimate, and its Page engagement rate counts clicks as interactions too. X shows impressions on every post. These aren't the same event, so the same 3% means different things on each.
- Some benchmarks divide by followers. That gives far smaller numbers than dividing by views or reach. Check what a benchmark divides by before you compare yourself with it.
For a rough guide, Buffer's 2026 engagement report divides engagements by reach and puts the 2025 median at about:
- LinkedIn: 6.1%
- Threads: 3.6%
- X: between 2.5% and 2.8% (the report gives both)
Those are medians for posts published through Buffer, and Buffer says itself that the numbers aren't directly comparable across networks. The comparison that matters is a post against your own median on the same network.
What to track instead: one set that works everywhere
Pick metrics every network can give you, then add engagement rate only where views exist.
| Goal | Track | Works on |
|---|---|---|
| Conversation | Replies per post | All five |
| Reach beyond your followers | Reposts and quotes per post, or reposts divided by followers | All five |
| Audience | Follower growth rate: new followers / followers at the start x 100 | All five |
| Attention | Engagement rate against your own median | Threads, X, LinkedIn |
| Traffic | Visits from UTM-tagged links | All five, read in your site analytics |
| Sign-ups or sales | Conversions from those visits | All five, read in your site analytics |
Replies are the most useful single number. Every network reports them, they take more effort than a quick tap, and a post that starts a conversation is usually the one worth repeating. We've written about why replies grow an audience separately.
Follower count on its own is context, not a goal. Watch the growth rate, because 200 new followers means a lot more at 1,000 than at 50,000.
Tag your links so every click counts the same way
Native click counts are patchy: X reports them for your own posts for 30 days, Threads only for the whole account, LinkedIn only on Page posts, and Bluesky and Mastodon not at all. UTM parameters fix that, because your own site analytics does the counting.
Add the same three tags to every link you post:
https://example.com/guide?utm_source=bluesky&utm_medium=social&utm_campaign=november-guide- utm_source is the network: threads, x, linkedin, bluesky, mastodon.
- utm_medium is always social, so these visits group together.
- utm_campaign names the post or series.
Keep them lowercase and spelled the same way every time, or your analytics will split one network into two rows. Then your click-through and conversion numbers come from one place, measured one way, for every network.
A review routine that takes 20 minutes a week
Weekly, 15 to 20 minutes. For each network, sort last week's posts by replies. Look at the top three and the bottom three, and write down one thing to repeat and one to drop. Then check that week's UTM visits by source.
Monthly, 30 minutes.
- Follower growth rate on each network.
- Engagement rate trend on Threads, X and LinkedIn, against your own median.
- Visits and sign-ups by source from your site analytics.
- This month's numbers, written down somewhere you keep. X stops giving tools click data after 30 days, so a monthly note is your long-term record.
Quarterly, an hour. Decide where your time goes. If one network brings most of the replies and sign-ups, give it more of your posts and adjust your schedule. For how many posts that is, our guide to how often to post covers each network.
Common mistakes
- Comparing engagement rate across networks. Compare each network with itself.
- Reading "0 views" on Bluesky or Mastodon as failure. Those networks don't count views, so judge those posts by replies and reposts.
- Treating LinkedIn impressions as exact. LinkedIn says they're an estimate. Watch the trend, not small differences.
- Judging a post on day one. Replies and reposts can keep arriving after the first day, and some LinkedIn Page metrics take up to 48 hours to show. Review last week's posts, not today's.
- Waiting too long to pull X clicks. After 30 days, tools can't read them any more.
- Tracking everything. Five or six numbers you act on beat thirty you only look at.
Where Shaflex fits
Shaflex puts posts from Threads, X, LinkedIn, Bluesky, Mastodon and your other connected networks in one table, with comments, reposts, likes, bookmarks, views, reach, link clicks and engagement rate, and you can sort it by any column. Sorting by comments is the weekly review above in one click.
It follows the same rules as this post. Where a network reports no views, as on Bluesky and Mastodon, the Views column reads 0 and the engagement rate is a relative score from likes, comments and reposts, not a share of views, so compare those posts only with each other. Shaflex doesn't read views for LinkedIn personal profile posts yet, so judge those by reactions, comments and reposts too.
Above the table, the Summary shows posts, impressions, followers, likes, comments and reposts, each with the change from the previous period. The Performance Overview charts follower growth, engagement rate and amplification rate (reposts divided by followers, which works on every network), plus an Engagement by Hour chart for your best posting hours.
Sources
Checked on November 16, 2026.
- Meta: Threads API insights
- X: Metrics (X API documentation)
- LinkedIn Help: Combined post analytics and Content analytics for your LinkedIn Page
- Bluesky: Show number of views for the post, open since January 7, 2025; post view definition (API lexicon)
- Mastodon: Status entity (API documentation)
- Buffer: State of Social Media Engagement 2026 (2025 data, posts published through Buffer, engagements divided by reach)
Questions readers ask
Which social media metrics matter most?
The ones every network reports and that track a goal: replies per post for conversation, reposts and quotes for reach beyond your followers, follower growth rate for the audience, and link clicks plus sign-ups (tagged with UTM parameters) for traffic. Engagement rate is useful too, but only on networks that count views, and only against your own past posts.
Can you see post views on Bluesky or Mastodon?
No. Bluesky returns likes, reposts, replies, quotes and bookmarks for each post but no view count, and a request to add one has been open on Bluesky's app repository since January 2025. Mastodon reports favourites, boosts, replies and, since version 4.5, quotes, and no views either. Compare posts there by replies and reposts instead of engagement rate.
What is a good engagement rate on social media?
Buffer's 2026 report, built on 2025 posts published through Buffer and dividing engagements by reach, puts the median at about 6.1% on LinkedIn, 3.6% on Threads and between 2.5% and 2.8% on X. Buffer itself says those numbers aren't directly comparable across networks. Use them as a rough guide and judge a post against your own median on the same network.
Can you compare engagement rate between networks?
Not fairly. Each network counts a view differently: Threads counts every time a post is played or displayed, LinkedIn calls its impression count an estimate, and Bluesky and Mastodon count no views at all. LinkedIn's own Page engagement rate even includes clicks. A 3% rate on X and a 3% rate on LinkedIn measure different things, so compare networks by replies, reposts and follower growth rate instead.
How do you track link clicks from Bluesky and Mastodon?
Add UTM parameters to the link, for example utm_source=bluesky and utm_medium=social, and read the visits and sign-ups in your own site analytics. Bluesky and Mastodon don't report link clicks, X only reports them for your own posts, and Threads reports clicks for the whole account, so UTM tags are the one click count that works the same everywhere.
How long do networks keep post analytics?
It varies. Through X's API, link clicks, profile clicks and other private metrics are only available for posts from the last 30 days. Threads' API returns insights back to April 13, 2024. If you want a long record, write your numbers down or export them once a month instead of relying on any one network to keep them.

