Most bad decisions with data start before anyone opens a spreadsheet. They start with a fuzzy question.

Three parts of a question data can answer

  1. Who or whatThe group you care about. "First-time customers", not "customers".
  2. A measureSomething you can count. "Repeat orders within 60 days", not "loyalty".
  3. A comparisonAgainst what? Last year, a target, another team.

Try it: flip each card

Each card shows a vague worry. Flip it to see an answerable version.

Vague

"Sales feel slow."

Answerable

Did weekly online sales between March and May fall compared with the same weeks last year?

Vague

"Students are struggling."

Answerable

What share of students submitted the Week 3 task on time, compared with Week 1?

Vague

"The new tool isn't working."

Answerable

How long did it take to close a support ticket in the month before and the month after the change?

An approximate answer to the right question is worth far more than a precise answer to the wrong one.John Tukey, statistician
What if I cannot get the data?

Then you have learned something useful early. Either find a measure you can get that is close enough, or decide that this question will be answered by judgement, and say so openly.

Designer's note

Flip cards ask learners to predict before they see the answer. That small moment of effort makes the answer stick far better than reading it straight away.