How to read a simulated number

These are not forecasts.

A synthetic population is more agreeable than a real one. It reads what is in front of it and does not get interrupted the way people do. So when Prefigure says 41% reached value, the useful claim is not 41. It is that roughly two in five got there, and that the losses are where the report says they are.

Four ways to use it well

  1. Compare, do not forecast. Run version A, change one thing, run version B. The gap is the finding.
  2. Read the quotes before the chart. The explanation of why holds up far better than the size of the number.
  3. Trust behavior over stated intent. Where someone left is evidence. Whether they say they would pay is not.
  4. Check the biggest prediction against reality. Ship the cheapest fix to real users and see whether the direction held.

What each number is worth

Where people leftHighCounted from the behavior log, not asked for.
Reached the valueHighA session either got there or it did not.
Came back on day 1, 3, 7MediumEach day is its own choice, made from how the last visit went and whether a reminder can reach them. Real boredom is still under-modeled.
Who the groups areMediumBuilt from how people in each of your target groups typically behave: why they come, what they expect and what makes them stop. Typical people of each kind, not anyone in particular.
What they think it doesMediumEach group says it in their own words after the first screens. Good at spotting confusion, imprecise as a score.
PaidLowCounted only when your screens include payment, through a simulated store sheet or card form. Real money behaves differently: rank versions with it, never budget against it.