Simulations

A simulation is one run: a set of personas, a target, and a goal.

Targets#

Fraser can run against:

• Live URLs: landing pages, pricing pages, signup and onboarding flows.
• Product flows: a sequence of steps rather than a single page.
• Screenshots and prototypes: useful before anything is built.
• Android application packages (APK files): see Testing Android APKs.
• iOS builds (beta). Run personas against an iOS app the same way as Android; upload a build that reaches the flow you care about, with a test account if it's behind auth.

All targets must be yours to test. See Authorization & limits.

Goals#

The goal is the highest-leverage setting in the whole product. It should describe an outcome the persona wants, not an instruction to the tool.

Good:

Work out whether this supports SSO on the team plan,
and how much that costs.

Weak:

Review the pricing page.

The first produces findings about whether the answer was findable. The second produces opinions about a page.

Scope#

Scope bounds how far personas may wander, which paths are in play, and where a run should stop. Narrow scope gives sharper findings; wide scope surfaces surprises. Start narrow.

Re-running#

Re-running the same personas and goal against a new build is how you tell whether a fix worked. Changing personas and target at once means you can't attribute the difference to either.

Treat a simulation config as something you keep, not something you rewrite each time.

Interpreting a run#

A single run is a sample, not a verdict. Consistent friction across personas is signal. One persona struggling where others sail through is usually about that persona's lens, which is still useful, especially for accessibility.

Next: Findings & reports.