Threads Analytics: Which Metrics Actually Matter

Meta expanded Threads Insights in 2025. You can tap into the Interactions figure to see likes, replies, quotes and reposts separately, and into Followers to see follower growth, top cities and countries, and age range and gender. A chart tracks views and interactions over 7 to 90 days, and Insights shows where people encountered your posts, including on Instagram and Facebook, where eligible Threads posts can now be recommended.

In September 2026 Meta redesigned the panel again. According to Engadget, it now adds AI-generated summaries, marks each post's views, likes and reposts as higher, lower or typical compared with your recent posts, and shows which communities drive engagement and which days and times reach your audience best.

Meta's developer documentation adds one detail that also applies to the API: follower demographics need at least 100 followers.

Link clicks. Meta added link-visit counts for bio and post links in May 2025, but Engadget reported in September 2026 that they were removed from the in-app Insights, with Meta saying only that it is exploring the feature. If you need click data, put UTM parameters on your links and read them in your own web analytics.

Saves. We found no evidence that Threads shows authors how often a post was saved. Treat any advice built on "save rate" with suspicion.

Meta also does not publish what counts as a good result. Any benchmark you see comes from a third party.

Replies per 1,000 views. Divide a post's replies by its views and multiply by 1,000. It tells you whether a post started a conversation, independent of how far it travelled. Compare posts with each other, not with other accounts.

Engagement rate. Buffer, which publishes the most detailed public Threads benchmarks we know of, defines it as likes, reposts, replies and quotes divided by reach. Use the same definition every month so your numbers are comparable.

Net new followers per month. Follower growth lags behind your content, so monthly is the right resolution. A good month after a run of strong posts is cause and effect; a single spike after a viral post is not a trend.

Buffer's 2026 analysis of more than 52 million posts across platforms put the median Threads engagement rate at about 3.6% in 2025, down from about 4.4% in 2024. By format, video posts had the highest median at 5.55%, followed by images at 4.55%, text at 2.79% and link posts at 2.34%.

These numbers come from Buffer's own users, who are mostly businesses and creators scheduling posts, so they are a rough reference point rather than a target. If your account sits far below them, look at your content; if it sits above them, do not assume you have nothing left to improve.

Take an invented example: a career coach who posted 14 times last month. She lists each post with its views and replies and calculates replies per 1,000 views. Her three best posts all ask readers about a specific workplace situation ("What did your last manager do in your first week that actually helped?"). Her three weakest are announcements of a webinar.

Two decisions follow. More posts built around one concrete workplace question, and webinar announcements folded into a post that is useful on its own instead of standing alone. Twenty minutes of work, and next month's plan is better than any general advice.