Twitter Analytics: 5 Metrics That Drive Growth

X's native analytics dashboard shows a dozen numbers. Five of them tell you something you can act on. The rest are context. Here's which is which.

1. Engagement rate: the number that actually predicts resonance

Engagement rate is total interactions (likes, replies, reposts, and link clicks) divided by impressions. X shows this per-post in the analytics view. A post with 10,000 impressions and 50 interactions (0.5%) is doing something different than a post with 1,000 impressions and 80 interactions (8%), even though the first has more raw reach.

What to actually do with it: sort your last 30 days of posts by engagement rate, not by impressions. The posts at the top of that list, not the impressions list, are the ones worth studying for format and topic patterns.

2. Reply rate: the leading indicator most creators ignore

Replies require more effort from a reader than a like, so a high reply rate is a stronger signal of genuine resonance than likes alone. It's also weighted more heavily in how X's ranking system distributes a post in the first place, per how the X algorithm actually ranks content. A post that generates a real back-and-forth in the replies is doing more for your account's ongoing distribution than a post that gets quietly liked and scrolled past.

Practical read: if two posts have similar engagement rates but one has proportionally more replies, that's the format to repeat, not the one with more likes.

3. Profile visits from a post: the metric that connects content to follower growth

X's per-post analytics show profile visits attributed to that post. This is the closest thing to a direct line between a specific piece of content and new followers, since someone has to click through to your profile before they can follow you. A post can have strong engagement and low profile visits if it's the kind of post your existing audience likes but that doesn't signal "follow this account" to someone new. Threads and posts that establish a clear point of view tend to drive more profile visits than short reactive takes, even when the short takes get more raw likes.

4. Link click-through rate: the one metric worth ignoring for scheduled posts

If your goal is audience growth, link CTR matters far less than the other four, and pursuing it actively can hurt you: X's distribution algorithm suppresses posts containing links, since the platform wants attention to stay on X rather than route it elsewhere. Track link CTR only when you're deliberately trading reach for a specific off-platform action (a newsletter signup, a product page), and expect that trade-off to cost you distribution on that individual post.

5. Follower growth velocity: the lagging metric that confirms the other four

Follower count itself is a lagging indicator, by the time it moves, the content that drove the move has already been published for a while. Tracking the week-over-week rate of change (not the absolute number) is more useful than watching the raw count daily, because it shows you whether a recent change in strategy, cadence, or format is actually compounding or just holding steady.

A worked example: two posts, same impressions, different signal

Post A gets 10,000 impressions, 200 likes, 5 replies, and 40 profile visits. Post B gets 10,000 impressions, 90 likes, 35 replies, and 110 profile visits. Raw engagement (likes + replies) looks similar at a glance (205 vs 125), but Post B is doing more for the account: nearly 3x the profile visits and 7x the reply count on the same reach. If you only look at total interaction count or impressions, Post A looks like the winner. Once you weight reply rate and profile visits, Post B is the format worth repeating, since it's converting attention into potential followers, not just passive approval.

Track your own baseline, not an industry benchmark

Published "average X engagement rate" benchmarks vary wildly between sources and rarely control for account size or niche, which makes them close to useless for an individual decision. What's actually useful is your own trailing 30-day average engagement rate as a baseline, then flagging anything that beats it by 2x or more as worth a closer look. This turns "is 3% good?" (unanswerable without a lot of context) into "is this post better than my last 30 days?" (answerable from your own dashboard in under a minute).

What to stop checking daily

Raw impressions and raw follower count, checked daily, mostly produce anxiety rather than decisions. Impressions swing with factors outside your control (time of day, unrelated trending topics competing for attention) and a single day's number rarely changes what you'd do differently. A weekly review of the five metrics above produces more useful signal than a daily check of the two vaguest numbers on the dashboard.

Why raw impressions is the most misleading number on the dashboard

Impressions count every time a post loads in someone's timeline, including scrolls-past with zero attention. A post can rack up impressions from being shown widely for a few seconds each, with almost nobody actually reading it, while a post with a tenth of the impressions but a genuinely engaged audience does more for the account. Impressions is a reach number, not an interest number. Treating it as the primary success metric is the single most common analytics mistake creators make, because it's the biggest, most visible number on the dashboard and the easiest to check obsessively.

The fix isn't to ignore impressions entirely, it's context for the other four metrics, not a target in its own right. A 5% engagement rate on 2,000 impressions and a 5% engagement rate on 20,000 impressions are the same quality signal at different scales; only the second one is also a bigger reach win.

Frequently asked questions

Where do I find these numbers in X's dashboard?

Each post has a small analytics icon (a bar-chart symbol) below it when you view your own posts, which opens per-post impressions, engagement, and profile-visit data. Account-level trends are under the main Analytics section of X's settings.

What's a "good" engagement rate?

It varies enormously by niche and account size, so compare a post against your own recent average rather than an external benchmark. A post at 2x your own typical engagement rate is a real outlier worth studying, regardless of what the absolute percentage is.

Does X show analytics for other accounts, not just mine?

No, per-post analytics (impressions, engagement rate, profile visits) are only visible to the account owner. For competitor or cohort research, you can only observe public signals: likes, reply counts, and repost counts.

How often should I actually check analytics?

Weekly is enough for the five metrics above. Checking per-post analytics immediately after publishing mostly measures how fast a post is being distributed in its first hour, which is useful only if you're actively deciding whether to boost engagement on that specific post through your own replies.

Do these five metrics apply the same way to threads as single posts?

Mostly yes, with one difference: a thread's analytics are reported per-tweet within the thread, so the first tweet's impressions are usually higher than later tweets in the same thread (some readers don't click "show this thread"). Compare a thread's opening post to your single-post baseline, and use the drop-off between tweet 1 and tweet 2 as a signal of whether the hook earned a read-through.