A client report is only as good as its weakest number. Here's a structure that leads with what changed, uses only figures the networks actually report, and says plainly where a figure doesn't exist.
Clients don't read reports for the numbers. They read them to find out whether things are going well and what you'll do next. The numbers are the evidence, which is why a single wrong one costs more than it seems: once a client catches one, they stop trusting the rest.
Start with what changed and why
Open with three or four sentences a client could forward to their boss: what happened this period, the one or two things that drove it, and what you're doing next. Everything after that supports those sentences.
A structure that works
- Summary. What changed, why, and what's next.
- Results against last period. A small table per network, each figure with its change.
- What worked. The three best posts, with a line on why each worked.
- What didn't. One or two honest lines. Clients notice when this is missing.
- Next period. What you'll test or change.
- Notes. What each figure means, and which ones a network doesn't report.
Use only figures the networks report
| Network | Per post | Per account |
|---|---|---|
| Views, reach, reactions, comments, shares, clicks | Page views, reach, post engagements, followers | |
| Views, reach, likes, comments, shares, saves | Reach, new followers, followers | |
| TikTok | Views, likes, comments, shares (public videos) | Followers, total likes, videos |
Meta no longer reports post impressions, Page fans or Instagram profile views; there's more in what Meta actually reports now. TikTok's API doesn't report reach or watch time; see what TikTok's API tells you.
Three mistakes that undermine a report
Adding up daily reach and calling it reach. Daily reach counts the people reached that day. Seven days added together counts anyone reached on two days twice. If you show the sum, call it the sum of daily reach.
Comparing across a metric change. When a network stops reporting a figure, last year's number has no counterpart this year. Putting both on one chart shows a collapse that's really a change in what the network reports.
Leaving gaps blank. A blank cell reads as zero, and zero is a result. Write "not reported by TikTok" or "not available before March" instead.
How often to send it
Monthly suits most clients: long enough for patterns, short enough to act on. Weekly makes sense during a launch or a paid campaign. Quarterly works as a review on top of monthly reports, not instead of them.
Building it without copying numbers by hand
FlashPosts keeps each connected account's posts and figures, imported from the networks with their history, and reports them per account, per client and per network, including when each audience is most active. Send the report as a PDF with your agency's logo every week or month, or connect Looker Studio and build your own. Each figure a network doesn't report is labelled rather than estimated.
Frequently asked questions
What should a monthly social media report include?
A short summary of what changed and why, results against the previous month for each network, the best and weakest posts with a reason for each, the plan for next month, and notes on what each figure means.
Can I still report impressions for Facebook and Instagram?
No. Meta stopped reporting impressions for posts. Use views, which count how many times a post was shown or played, and reach, which counts the people who saw it.
How do you compare reach across Facebook, Instagram and TikTok?
You can't fully, because TikTok's API doesn't report reach. Compare views across networks instead, and say in the report that TikTok reach isn't reported.