Social media analytics: what to track every week

TL;DR

• Check five numbers every week: reach, saves, shares, comments with profile visits, and engagement per view (interactions divided by reach).
• Compare each post with your own previous period, never with industry benchmarks, and remember that reach does not add up across days.
• Turn the review into one change for next week: repeat the winning topic in a new format or move the posting time.

Most people open their analytics, look at the follower count, feel good or bad about it, and close the tab. Nothing changes in what they post next week.

Analytics are only useful when they change a decision. That needs fewer numbers, read the same way every week, and compared against your own past instead of somebody else's benchmark.

Here are the numbers worth that weekly look, what each one actually tells you, and the routine that turns them into next week's posts.

Reach: did the post get shown?

Reach is how many different accounts saw the post at least once. It answers the first question about any post: did the platform give it a chance?

Low reach with good engagement from the people who did see it means the post was right and the distribution was not: try a different time, a stronger first slide, or a format the platform is pushing. High reach with nothing else means people saw it and kept scrolling.

One warning that catches almost everyone: reach does not add up across days. The same person seeing your posts on Monday and on Tuesday is one account, not two. A tool that sums daily reach into a weekly number is inflating it.

Saves: is it worth coming back to?

A save is somebody telling Instagram "I want this later." It is the strongest signal of usefulness you can get, and on Instagram it weighs heavily in what gets shown to more people.

Carousels with steps, checklists, templates and references collect saves. If a post had average reach but twice your usual saves, that is the topic and format to repeat.

Shares: is it worth sending to someone?

Shares are reach you did not have to earn yourself: someone sent your post to a friend or put it in their story. A post that gets shared is one that made someone look smart or helpful for passing it on.

Comments and profile visits: did it start something?

Comments tell you the post provoked a reaction. Profile visits tell you it made someone curious about who wrote it. For a business, profile visits are often the most underrated number on the screen, because a visit is the step right before a follow, a click on the link in bio, or a DM.

Engagement per view: the number that lets you compare

Raw counts favor the posts that happened to get more reach. To compare a post seen by 800 people with one seen by 8,000, divide the interactions (likes, comments, saves, shares) by the reach. That ratio is the fairest single answer to "which post worked better?"

The weekly routine, in fifteen minutes

1. Sort last week's posts by engagement per view. Look at the top two and the bottom two, not the whole list.
2. For the top two, write down why. Topic, format, first slide, time posted. One line each.
3. Compare with the previous period, not with last year and not with a competitor. Is reach trending up or down? Are saves?
4. Check when your audience responded. Your own best hours beat any published chart. We wrote a whole post on finding your best time to post.
5. Change one thing next week. Repeat the winning topic in a new format, or move the posting time. One change, so you know what caused the result.

What to ignore

• Follower count, week to week. It moves slowly and lags behind everything else. Look at it monthly.
• Likes on their own. They are the cheapest interaction and the weakest signal.
• Industry benchmarks. "The average engagement rate is 1.2%" tells you nothing about your audience, your niche or your format.

Where Zynta Flow fits

Zynta Flow connects to Instagram, LinkedIn and TikTok through their official APIs and imports up to 90 days of history on the first connection, so you start with a baseline instead of an empty chart. The first import is free. After that it keeps a daily snapshot of each connected account.

The analytics screen shows reach, saves, shares, comments, profile visits and engagement per view for every post, compares the period with the previous one, and finds the hours your audience responds. The dashboard turns that into suggestions, like a post worth repeating or a week with nothing scheduled. When you want it in words, Zynta writes an analysis of the period, what went up, what went down and what to try next, or a full report you can share with a client.

And because the same app creates and schedules the posts, the routine above ends where it should: with next week's post already on the calendar.

Frequently asked questions

Which social media metrics matter most?

Reach shows whether the post was shown, saves show it was useful, shares show it was worth passing on, and profile visits show it made someone curious. Engagement per view lets you compare posts with different reach.

How do you calculate engagement per view?

Add the likes, comments, saves and shares of a post and divide by its reach. The ratio lets you compare a post seen by 800 people with one seen by 8,000.

Can I add up daily reach to get weekly reach?

No. Reach counts unique accounts, and the same person seeing your posts on two days is one account. Summing daily reach inflates the number.

How often should I check social media analytics?

Once a week for posts, in about 15 minutes, and once a month for follower growth, which moves slowly.

Does Zynta Flow import past posts?

Yes. When you connect Instagram, LinkedIn or TikTok, Zynta Flow imports up to 90 days of history, and the first import is free.