Map each role to its own aha moment

Capturing the role is worthless unless each role has a different destination, and most teams skip this step because it is the hard one: you have to decide, per role, what first outcome counts as value. The distinction that matters is between the aha moment and the activation event. The aha moment is the emotional realisation that the product is worth keeping; the activation event is the measurable proxy you can build a route towards. You design the route to the event because you can instrument it, but you choose the event because it best approximates the feeling. This is the same discipline that underpins choosing a north-star metric: pick the one measurable thing that best stands in for real value, then build everything towards it.

For every role you support, write the activation event as a concrete, single sentence with a verifiable outcome: "the admin has invited at least one teammate and connected one data source", "the analyst has built one chart from their own data", "the operator has completed one real task end to end". These are not the same milestone, and forcing them through one shared checklist guarantees that at least two of your three roles are doing busywork that does not move them towards their own value.

Once each role has its event, work backwards to the fewest real steps that reach it. "Honest" matters here: a fake-data demo that lights up a green checkmark is not activation, because the aha moment is emotional and a person feels the difference between their own data and a sandbox. Personalised, role-specific paths are what produce the 30 to 50 percent activation lift the case studies report, and the reason is mechanical, not magical: you have deleted every step that was value for some other role and kept only the steps that are value for this one. Time to value collapses because the path is now the user's path, not the average of all paths. Recent data backs the size of that collapse, with teams streamlining onboarding around the user's own goal reporting time-to-first-value cut by as much as 45 percent.