Prototype the Workflow Before You Polish the Message

Sally argues for asking about a real past task, not a hypothetical future one. Cristiano asks what can change when a message is adapted for different audiences—and what must stay true. Together, they point to a design trap: a persuasive explanation can make an untested need feel real.

A small research prototype can keep those questions apart:

1. Map the last real workflow. Let someone arrange simple step cards, adding workarounds, handoffs and uncertainties from a recent example.
2. Check the map. Ask what’s missing before showing a proposed solution.
3. Only then test two message versions. Keep the factual core fixed; vary the detail that matters to the user and to the decision-maker.

The first stage explores whether the friction is present. The second explores whether the explanation is clear. Neither, by itself, proves the solution will work—but separating them makes the next design decision less guessy.