Blizsocials

INSIGHTS

Reading your own analytics without lying to yourself

Which numbers actually predict revenue, which ones only feel good, and the monthly report we send clients.

Four Blizsocials team members in a boardroom reviewing printed coverage and reports laid out in front of them

Every platform hands you a dashboard designed to make you feel good about using the platform. That is not a conspiracy, it is just what the numbers at the top of the screen are for. The problem is that most brands then report those numbers to themselves every month, feel encouraged, and change nothing. Here is how to read the same screen in a way that actually moves a decision.

The short version

• Reach, impressions and follower count are the three easiest numbers to fool yourself with.

• Saves and shares predict revenue better than likes do.

• Always compare against your own last month, never against a stranger's screenshot.

• If a number cannot change a decision, stop reporting it.

The three numbers that flatter you

Reach, impressions and follower count. All three go up when you post more, all three go up when one post gets lucky, and none of them tell you whether a single person moved closer to buying. They are worth watching as a trend over months. They are worth almost nothing as a monthly score.

The reason they persist is that they are the easiest numbers to put in a slide, and a rising line is a pleasant thing to show a boss. Which is fine, until the quarter ends and somebody asks what the line bought.

The ones that actually predict money

Saves and shares, first. A save is somebody saying "I will need this later", and a share is somebody staking a bit of their own reputation on you. Both are far closer to intent than a like, which costs nothing and means nearly nothing.

Then profile visits and link taps, which measure the thing most accounts never check: whether the content made anybody curious enough to go looking. Then DMs and comments that are questions rather than compliments. And then the only one that finally matters, which is enquiries, and which almost nobody bothers to connect back to the post that caused them.

You do not need attribution software for that last one. You need to ask people where they found you and write the answer down.

A number with nothing to compare it to is not a result. It is a rumour with a decimal point.

Compare against yourself, not a screenshot

The fastest way to make a good month look bad is to compare it to somebody else's carousel on LinkedIn. Different industry, different audience size, different amount of money behind it, and quite possibly a number that has been quietly rounded up.

Compare this month to your own last month, and this quarter to the same quarter last year, because most Nigerian businesses have a season and yours is probably in one right now. A December that beats November means less than a December that beats last December.

The five minute monthly review

Three questions, in this order, and nothing else.

• What moved, and can we say why? Not "engagement was up", but which posts, and what those posts had in common.

• What did not move, and did we expect it to? A flat month you predicted is information. A flat month that surprised you is a warning.

• What changes next month because of the two answers above?

If the third question has no answer, the review was a status update, not a review. This is the single most common failure we see in reports written by other agencies, and it is why we insist a report ends in a decision.

The honesty tests

Two, and both are uncomfortable. First: if a number cannot change a decision, stop reporting it. It is decoration, and decoration in a report is how a bad quarter hides.

Second: write down what you expect before you open the dashboard. Then open it. The gap between what you expected and what happened is the only genuinely new information on the screen, and it disappears the moment you look at the numbers first and decide afterwards that you knew all along.

What we report, and why it is short

Our monthly reports run to a page or so. What moved, what did not, what changes, and the numbers that support those three answers. Not a screenshot of a dashboard the client can open themselves, and not forty slides of context nobody reads to the end of.

When a month goes badly it says so, with the reason and the fix, because an agency that only sends good months is managing the client rather than the account.