How to build a social media report somebody actually reads

Forty two slides went out on the second, branded, every chart the tool could export. On the fourth the client replied with one line asking something the deck did not answer, and the answer took an unbilled afternoon. The report was thorough. It was also read for about ninety seconds, and the only part anybody remembered was the sentence at the top.

This is how to build a social media report that gets read and acted on. It takes about an hour a month once set up, and most of the work is deciding what to leave out.

TL;DR

Step one: write the first line before opening any tool

One sentence at the top saying what happened and what you are doing about it. Write it first, from what you already know, and let it decide which numbers are worth pulling.

Done the other way round, the report becomes whatever the export contained, and a report shaped by an export is a report nobody can act on. This single reversal is the difference between an hour and an afternoon.

Forty slides is not thoroughness, it is an inability to decide which three numbers mattered.

Step two: lead with the commitment, not the performance

The first section is what was agreed and whether it happened. Posts promised against posts published, per person or per account. It is the row clients remember and the one most reporting tools cannot fill, because the commitment lived in a contract rather than inside a connected account.

It also reframes every difficult month. A quarter with weak engagement and a held cadence is a content problem you can work on. A quarter with a missed cadence is a different conversation entirely, and pretending the two are the same is how a retainer ends without anybody understanding why.

Step three: three numbers, each with a trend

Three, not twelve. A number without a comparison is decoration, so each one carries the previous period beside it and nothing else.

1. Posts published. The only figure entirely within your control, and the one that predicts the others.
2. Audience over ninety days. Not week on week. Follower counts are noise at weekly resolution on any account under a few thousand.
3. Engagement, one definition, stated. Computed identically every month, with the formula in the footer.

The formula matters more than the value. Two months computed differently is a trend line describing your own method, and the fix is one line of text. It is set out in how to calculate engagement rate.

Step four: one thing that worked, one that did not

Two examples, with the actual posts, and a sentence each on why. This is the section that takes judgement and it is the reason the client is paying somebody rather than reading a dashboard.

Including the failure is not modesty, it is credibility. A report where everything worked is a report nobody believes, and the month something genuinely goes wrong you will have no vocabulary for it because you have never used any.

Step five: label every estimate

Any report containing competitors contains modelled numbers, because a competitor has connected nothing. That is normal and acceptable. What is not acceptable is a slide where a measured figure and an estimate share a column and a font.

A footnote on every modelled row costs one line. It protects you the day a client checks a figure against their rival's own reporting, and it makes the measured rows more credible by contrast. Say what you are not measuring too: reach on personal profiles, for instance, and why.

Step six: question the monthly rhythm

The monthly report exists because agencies invoice monthly. Nothing about social publishing runs on that rhythm, which is why the document always feels either premature or overdue.

• A live view, always on. The client looks when they are curious, which is when the answer is worth most, and the fourth of the month stops generating email.
• A short note when something changes. Two paragraphs the day a pattern breaks beats forty slides three weeks later.
• A quarterly review with judgement in it. The meeting worth preparing for, and the one clients renew on.

Agencies resist this because the deck is visible proof of work. Proof is only needed when the results are not obvious, and a live view makes results obvious faster than a document. The version of this for personal branding retainers is in client reporting for personal branding.

Step seven: keep the raw record, separately

Whatever the report looks like, keep the underlying readings with their dates somewhere that is yours. Reports get rebuilt, tools get switched, and a client relationship ends.

An agency that can open eighteen months of a former client's publishing record walks into the re-pitch with evidence nobody else has, and that only exists if the data was never rented from a connection that ended with the contract. Which products keep it on your side is compared in best social media reporting tools for clients.

What to hand over when a client leaves

Churn is the moment reporting choices get graded. A client who leaves with nothing tells the next agency you were opaque, and a client who leaves with a clean record often comes back, because the comparison they run afterwards is against your baseline.

• The raw publishing record. Dates, formats and counts, in a file they can open without your login.
• The definitions. How every rate was computed, so the next agency cannot quietly redefine improvement.
• What you would do next. One page. It costs an hour and it is the cheapest re-pitch you will ever write.

Preparing for churn feels like expecting it. It is the opposite: a clean handover is the cheapest reputation available, and the agency that gives one is the one named when somebody at their next company asks for a recommendation.

Getting it down to an hour

The reason reports take an afternoon is almost never the analysis. It is assembly: finding numbers, rebuilding the same chart, screenshotting a dashboard that will look identical next month.

Two of those five deserve your time and three of them do not. If the split in your month is the other way round, the problem is the pipeline rather than the report, and it is worth an afternoon once to fix rather than an afternoon every month to endure.

A final note on who the report is for. Written for the person paying, it stays short and leads with the commitment. Written to protect the agency, it grows every month, because each awkward question adds a slide and no slide is ever removed. If your deck has grown for six months straight, that is what happened, and the fix is to delete rather than to add.

The template is the cheap part. Who the report is for, what to leave out however available it is, and how to report a bad month, is in social media reporting.

How Groowth handles this

It fills step two and step seven. Paste public handles, set the cadence each account agreed to, and it reads the profiles daily: what was published and when, the follower count on the day, and the public engagement. The commitment sits next to the reality on the same card.

One organisation per client, with a read only seat you hand over, so the client can look on the fourth without emailing you. It does not build a branded document and reports no paid, web or email channel, so for a multi channel contract it sits beside a reporting platform rather than replacing one. Three accounts free with no card.

Frequently asked questions