Audience study blocks

Audience blocks evaluate your screens from a selected audience's point of view — every finding says which audiences it came from, and every audience block also returns one result per audience. All five accept freeform Additional notes to steer the AI toward what you care about.

First impression

Simulates the first seconds with a screen: what each audience thinks this product is, what stands out, what confuses. No configuration beyond optional notes — the point is the unprimed reaction.

Persona reaction

A deeper read than first impression: each audience reacts to the screens as a whole — appeal, relevance, objections, and whether the value proposition lands for them. Results come back per audience with a sentiment and summary, so two audiences disagreeing is a finding, not noise.

Task walkthrough

Gives the audience a task and walks it through your screens. Configuration:

• Task or goal (required) — what the user is trying to accomplish.
• Success criteria (optional) — how you'd know they succeeded.

Each audience's result predicts the path they'd take and whether they'd succeed. Findings call out where the walkthrough stalls: unclear next steps, missing affordances, wording that doesn't match the audience's vocabulary.

First click

The classic first-click test, simulated: where would this audience click first to do X?

• Task question (required) — the "where would you click to…" prompt.
• Screen (optional) — which screen the question is asked against.
• Expected target (optional) — what you consider the right answer.

Each audience's result names its predicted first click and the targets competing for it.

Design comparison

Puts two or more screen variants head-to-head for the selected audiences. Configuration:

• Comparison goal (required) — what "better" means for this comparison.
• Screens to compare (required, at least two) — the variants, picked from the study's screens.

Each audience's result names the option it would prefer and the tradeoffs behind that preference.