Strategy
If the number changes nothing, it's reporting
Most monthly reports are accountability ceremonies. They get read, approved, and change no decision at all. That isn't measurement.
Reach, impressions, engagements, follower growth, clicks. An average monthly report carries between twelve and twenty figures. Almost none of them change what gets done next month.
That isn’t measurement. It’s activity documentation, and it exists to prove work happened, not to decide anything.
The test
For every metric you’re about to include, ask one question: what would I do differently if this number doubled? And if it halved?
If the answer to both is “nothing,” the metric isn’t strategic. It can stay as context, but it shouldn’t hold the center of the report or the conversation.
Most reach metrics don’t survive this test. If impressions rise thirty percent, almost nothing changes. If they drop thirty percent, also nothing. It gets noted and everyone moves on.
A metric is only strategic when it changes a decision.
Each intent has its own metric
The original error is measuring everything the same way. Communication can pursue three different things, and each is verified differently:
- Recognition. That people know you exist. Verified through reach against the target audience, recall and branded search. Here reach is the correct metric.
- Education. That they understand something they didn’t before. Verified through depth of consumption, return visits and questions that stop appearing in the sales process.
- Action. That they do something. Verified through conversion, registration quality and close rate.
When you ask an action campaign for reach, or a recognition campaign for conversion, the report will say something failed even when the campaign did exactly its job.
The most expensive vanity metric
It isn’t reach. It’s the lead count.
A lead is not a result. It’s a halfway intervention. The result sits two or three steps further on: contact established, appointment booked, enrollment completed. A report celebrating four hundred registrations without saying how many were reachable is hiding the number that decides.
The correction is simple and almost nobody applies it: report the gross figure and the useful figure together. If a hundred and twenty of four hundred registrations were contactable, that’s the number. The other two hundred eighty are spend.
Perfect attribution doesn’t exist
With platform fragmentation and privacy restrictions, no system attributes exactly. Chasing that precision consumes time that pays better elsewhere.
What does work is consistent attribution. If you measure the same way every month against the same criteria, even imperfect ones, you can compare against yourself and detect real change. Switching attribution models mid-campaign destroys the entire historical series.