Twitter analytics changed shape twice over: the platform became X, and the account-level dashboard that used to be free moved behind X Premium. So the first thing a creator needs in 2026 is a clear map of what you can still see for free, what you now have to pay for, and which of the numbers in either view are worth your attention. This guide gives that map. Free accounts still get per-post analytics in the mobile app — tap a post, open the menu, choose View analytics, and you get impressions, engagements, profile visits, new follows, and the link clicks that, oddly, the paid dashboard doesn't surface. The full account-level view — trends over time, engagement rate, audience, your most active hours — now requires Premium. The guide defines every metric precisely (impression versus reach, what counts as an engagement, how engagement rate is actually calculated), flags the gaps the native tools don't cover (reach for organic posts, sentiment, share of voice, who reposted you), and is honest that published engagement benchmarks for X are all over the map — one study's 0.015% and another's 2% are measuring different things — so your own trend line is the only benchmark that reliably means something. Then it turns to the part every analytics guide skips: a dashboard is a diagnostic, not an outcome. Knowing which post worked is worthless unless you make more of what worked, and faster than the feed forgets — and on a platform where the average post earns a couple dozen likes and impressions have been drifting down year over year, the winning move isn't a prettier chart, it's turning each signal into more content, on X and everywhere else the audience lives. That production gap is where a content engine like Kompozy fits.
Two things happened to Twitter analytics, and they stacked. Twitter became X, and the account-level analytics dashboard that used to be free for everyone moved behind the X Premium subscription. The practical result in 2026 is a split: anyone can still see how a single post performed, for free, in the mobile app — but the historical, account-wide view (trends over time, engagement rate, audience, your best posting hours) is now a paid feature. So before you optimize anything, you need a clean map of what lives on which side of that line.
This guide is that map, and then the part most analytics guides skip. It defines every metric X reports precisely — because impressions, reach, and engagement are routinely conflated — tells you exactly where to find the free view and the paid one, is honest about what the native tools simply don't cover, and warns you off the published engagement benchmarks that disagree by a factor of a hundred. Then it makes the argument that matters: a dashboard is a diagnostic, not an outcome. Reading which post worked is the easy half; making more of what worked, fast enough that the feed hasn't moved on, is where the actual leverage is.
Start with the access split, because it's the single biggest change and the one most out-of-date guides get wrong. Free accounts retain post-level analytics in the X mobile app: tap any of your posts, open the three-dot menu, and choose View analytics. That view gives you, for that one post, impressions, total engagements, detail expands, profile visits, and new follows — and, notably, link clicks, which the paid dashboard does not surface. It is genuinely useful for post-by-post learning, and it costs nothing.
What now requires X Premium is the account-level dashboard — the view that aggregates across posts and time. That's where you get engagement rate, follower growth trends, side-by-side metric comparisons, audience data, and your most active hours. You reach it on desktop by opening Premium in the left panel and choosing Analytics (or going directly to x.com/i/account_analytics), and in the app by tapping your profile photo, then Premium, then Analytics. Premium is a paid tier held by a few million accounts — a small fraction of the user base — so for most people the honest situation is: free gets you per-post numbers, and the trend view is a subscription decision. The social media analytics glossary entry covers the general vocabulary this all sits inside.
Analytics is only useful if you know what each number actually counts, and on X several of them are routinely confused. Here are the definitions that matter, stated carefully.
Impressions count how many times your post was displayed, full stop — if one person scrolls past it twice, that's two impressions. Reach counts unique accounts that saw it. The distinction matters because a high impression count can hide a small audience seeing your post repeatedly. X's native analytics reports impressions but does not surface reach for organic posts, so true reach is something you either infer or pull from a third-party tool. For day-to-day work, impressions is the headline distribution number you have.
An engagement on X is any deliberate interaction with the post: likes, replies, reposts and quote posts, bookmarks, link clicks, clicks on your handle or profile photo, detail expands, and follows earned off that post. Engagement rate is the one derived metric worth watching, and the standard formula is total engagements divided by total impressions, multiplied by 100. Dividing by impressions — the people who actually saw it — rather than by your follower count is what makes it an honest measure of how the content performed with the audience it reached, which is why it's the comparison the paid dashboard leads with. See the engagement rate definition for the cross-platform version of the formula.
Profile visits count clicks on your name or handle that open your profile — a leading indicator that a post made someone curious about you rather than just the tweet. Follower metrics in the dashboard break into new follows, unfollows, and net change over time, which is more useful than a single follower count because it tells you whether a given week actually grew the account. Mentions (times people tag your @handle) show in the Premium dashboard and in Notifications. Reposts are reported as a total count only — X does not natively tell you who reshared you.
For video posts, X reports video views (counted after a set amount of playback) and, in the richer views, a completion rate — the share of viewers who watched to the end. Completion rate is the more honest signal of whether the video held attention, since a view can be triggered by a few seconds of autoplay. If video is a real part of your strategy on X, watch completion over raw views.
Being clear about the gaps is more useful than pretending the native dashboard is complete. It isn't. Reach for organic posts isn't surfaced. Link clicks, oddly, appear only at the post level on mobile and not in the paid account dashboard. And there's a whole category the native tools don't touch at all: sentiment (are mentions positive or negative), share of voice (how your presence compares to competitors or a topic), the specific accounts reposting or mentioning you, and any competitive benchmarking. Those are listening-and-reporting jobs that require a third-party social-analytics tool — the category covered in best social media analytics tools for creators. Native X analytics is strong on your own posts and account trend, and blind to context.
You will find no shortage of articles quoting an "average X engagement rate," and they disagree wildly — one study reports something like 0.015%, another lands near 0.5%, another cites 1–2%. These aren't contradictions so much as different measurements: different sample sizes, different account tiers, different definitions of what counts as an engagement, and different denominators (impressions versus followers). The lesson isn't to find the "right" number — it's that a single external benchmark is close to meaningless for your account. The benchmark that reliably means something is your own trend line: is this month's engagement rate above or below your trailing average, and which posts moved it. Our social media engagement rates by industry guide walks through why two "official" figures can be ten times apart.
Two pieces of platform context are worth holding while you read your own numbers. Published benchmarks suggest the typical X post earns only a couple dozen likes and a handful of reposts, and that average impressions per post have been drifting down year over year — so low absolute numbers are the norm, not a sign you're failing, and the realistic goal is relative improvement and volume rather than a viral outlier. Treat those figures as directional, not gospel; they come from third-party studies, not an official X report.
You don't need a paid tool to run a disciplined loop. Weekly, open the post-level analytics on your best and worst posts and ask one question of each: what did the winner have that the loser didn't — the hook, the format, the time, the topic. If you have Premium, use the account dashboard for exactly two things: the engagement-rate trend (up or down versus your own baseline) and your most active hours (to inform scheduling, not to obsess over). Everything else in the dashboard is interesting and mostly not actionable. The point of the routine is to end each week with one concrete hypothesis about what to make more of — which is the handoff to the real work.
Here's the honest problem with every Twitter-analytics guide, including the first three-quarters of this one. Analytics tells you what already happened. It is a rear-view mirror, and a very good one — but no amount of reading it changes a single future number. The only thing that moves your metrics is publishing more of what the data says works, and doing it before the feed moves on. On a platform where the average post earns a couple dozen likes and impressions are drifting down, the gap between a creator who grows and one who stalls is rarely the quality of their dashboard. It's throughput: how reliably they turn last week's winning signal into this week's content.
That's the bottleneck analytics exposes but can't fix. You learn that your founder-to-camera clips outperform your text posts two-to-one, or that a particular angle on a topic spikes engagement — and then you have to actually make ten more of those, keep them on-brand, and also feed the TikTok, the Reels, the LinkedIn feed, and the newsletter where the rest of your audience lives. Reading the analytics is an afternoon. Acting on it at that cadence is the job, and it's where most creators' loops break.
Kompozy is the production half of that loop — the engine that takes what X analytics told you worked and turns it into more of it, everywhere, fast. The useful way to see the fit is as a diagnostic-to-supply handoff: your analytics names the winning format, angle, or hook; Kompozy manufactures the next wave of content in that shape from a single input, across 18 formats. If the data says talking-head clips win, you generate more Persona Shorts; if a quote-style post spiked, you spin Quote Graphics and Persona Tweets; if a topic landed, it becomes a Carousel, a Photo Post, a Blog recap, and a newsletter — not re-posted, but regenerated as native content for each surface.
Two things make the volume usable rather than noisy. First, everything stays on one brand: a written Persona Brief governs voice and rejects off-message copy, Gemini face-lock holds one presenter across a whole run, and HyperFrames renders brand-exact styling, so twenty pieces read as twenty takes from the same account. Second, Kompozy doesn't just generate — it schedules and publishes, fanning output across eight social platforms plus blog and email on Autopilot behind a per-post review gate, so acting on your X insight isn't ten manual re-uploads. The honest boundary: Kompozy is a content generation-and-publishing engine, not an analytics product — it won't replace your X dashboard or a dedicated listening tool, and it doesn't read your paid metrics back in. You still measure in X (and in the native views this guide maps); Kompozy is what you do with what you measure. For the X-specific growth side, see how to build a brand presence on X; for the cross-platform analytics mindset, Threads creator analytics and Facebook analytics for small business cover the same diagnostic-to-action loop on neighboring platforms.
Partly. Free accounts can still open per-post analytics in the X mobile app — tap the post, open the three-dot menu, and choose View analytics — which shows impressions, engagements, detail expands, profile visits, and new follows for that single post. The full account-level dashboard — trends over time, engagement rate, audience data, and your most active hours — now requires an X Premium subscription. So the split is: post-level metrics free, the historical dashboard paid.
Two paths. For a single post on mobile, tap the post, open the menu, and choose View analytics. For the account-level dashboard, you need X Premium: on desktop open Premium in the left panel and choose Analytics, or go to x.com/i/account_analytics; in the app, tap your profile photo, then Premium, then Analytics. The account dashboard is where you get engagement rate, follower trends, audience breakdowns, and posting-time data.
The standard formula is total engagements divided by total impressions, multiplied by 100. Engagements on X count likes, replies, reposts and quote posts, bookmarks, link clicks, clicks on your handle or profile photo, detail expands, and follows earned from the post. Dividing by impressions (people who actually saw it) rather than by followers gives a truer read of how the content performed with the audience it reached.
Impressions count how many times your post was displayed — one person scrolling past it twice generates two impressions. Reach counts unique accounts that saw it. X's native analytics reports impressions but does not surface reach for organic posts, so if you need true reach or deduplicated audience numbers you have to infer them or use a third-party tool. For most creators, impressions plus engagement rate is the workable native pair.
Several things. It doesn't surface reach for organic posts, link clicks are only visible at the post level (not in the paid dashboard), and it has no sentiment, no share of voice, no list of who reposted or mentioned you, and no competitive benchmarking. Those require a third-party social-analytics or listening tool. Native X analytics is strong for your own post and account trends and weak for context about how you compare or how people feel.
Twitter (X) analytics in 2026 is split: free accounts see per-post metrics in the mobile app (tap a post, open the menu, choose View analytics) — impressions, engagements, profile visits, new follows, and link clicks — while the full account-level dashboard with engagement rate, follower trends, and audience data now requires X Premium. Engagement rate is engagements divided by impressions times 100. The native tools don't surface reach for organic posts, sentiment, or who reposted you — those need a third-party tool.
Get started → · ← All guides · Compare Kompozy vs other tools