Someone in the meeting says “let us look at the data and decide”, and a dashboard goes up on the screen. Thirty minutes later what remains is a handful of numbers and a shared impression that traffic is lower than expected. Next week the same screen goes up and the same impression is exchanged.
This is not a tooling problem. A dashboard opened without a question produces observations, not decisions. Only the order needs to change: before opening anything, write down in one sentence what you are trying to decide.
A dashboard is a lookup tool, not an answer
What a dashboard shows you is everything that happens to be measurable. Open it without a question and you naturally follow whichever number stands out — and the number that stands out is usually the one that moved most. Nothing guarantees that is the number that matters to you this month.
Worse, observations make meetings longer. Everyone arrives having looked at a different screen and argues from a different figure, so the conversation drifts from “what shall we do” to “whose number is right”. When the question is fixed first, the screen to look at is fixed with it and that drift never starts.
Keep only the questions whose answer changes something
Not every curiosity is a good question. The ones worth a planner’s time pass three tests.
The first test does most of the work. If you would do the same thing whatever the answer turned out to be, there is no reason to look right now. That single sentence hands back a surprising amount of meeting time, and the data will still be there when it starts to matter.
The third test is the one most often skipped. A question with no date attached quietly becomes “let us look a bit more first”, and while it waits, so does the decision it was meant to serve.
Write the question out as a sentence
Questions that sound reasonable in your head reveal their vagueness once written down. The left column below is the form these usually arrive in; the right column is the same concern rewritten so it can produce a decision.
Notice that you can picture the shape of the answer for everything on the right. If you cannot imagine what the answer would look like, what you have is closer to a mood than a question — and that is a useful signal that it needs another pass.
Keep the questions you write rather than discarding them. As with much of what sits in the development workflow archive, the questions that keep coming back are the ones that reveal what the organisation actually cares about. If you would rather start from a documented picture of the site’s current state, the diagnostic report in our optimization program makes a reasonable starting point.
Next part
With the question settled, the screen to look at settles too. The next part is search data, starting with the combination that is most often misread — pages that get impressions but no clicks.