AI Prompts for Product Teams

Evaluate this product's onboarding flow against the Trust Equation: T = Usefulness × Predictability × Visibility. Score each variable from 0 to 10 and explain which factor is most likely to collapse trust for new users. Be specific about what you see and what is missing.

A user has just received an AI recommendation they didn't expect. Using the Trust Equation, identify which variable failed. Was it Usefulness (the recommendation wasn't relevant), Predictability (the behavior surprised them), or Visibility (they couldn't see why the AI did it)? Suggest one change per variable.

Review this copy and UI for an AI product. Score its Visibility. Does the user understand what the AI is doing and why? Give a specific score and rewrite any element that scores below 6 out of 10.

Apply the Trust Equation to a competitor's AI product. Which variable is their weakest? What would a product with a perfect score look like in this category?

Evaluate this feature against the Manipulation Matrix. Plot it on two axes: how much it benefits the user (low to high) and how much it benefits the company at the user's expense (low to high). Which quadrant does it land in? Is this ethical influence or a dark pattern?

This retention mechanic keeps users in the product. Analyze it using the Manipulation Matrix. Would the user thank you for it in one year, or would they feel trapped? Name the quadrant and explain your reasoning.

Review this notification strategy. For each notification type, determine whether it is persuasion (user-aligned) or manipulation (company-aligned at user expense). Rewrite any that cross the line.

Design an ethical version of a known dark pattern, such as confirm-shaming, the roach motel, or a hidden unsubscribe. Use the Manipulation Matrix to verify your redesign lands in the ethical quadrant.

Audit this AI product for transparency. Answer: Does the user know when they are talking to AI? Do they understand how the AI made its decision? Can they see what data was used? Score each dimension 1 to 10 and recommend one concrete fix per dimension.

This AI product surfaces recommendations. Does it tell users why? Write three versions of the explanation. One opaque (current), one translucent (some context), one transparent (full context). Which one builds the most trust without overwhelming?

Evaluate the error states of this AI product. When the AI fails or is uncertain, what does the user see? Score the transparency of failure. Rewrite the error message to be honest without being alarming.

A user asks why the AI did that. Write the ideal response for this product context. Specific, honest, and short enough to actually be read.

Evaluate this team's AI user experience practices against the 5-level maturity model. Level 1 Reactive, Level 2 Intentional, Level 3 Systematic, Level 4 Anticipatory, Level 5 Transformative. Which level are they at? What are three specific practices that would move them to the next level?

Review this product's approach to AI user research. Are they testing with real users in real contexts, or are they relying on metrics alone? What maturity level does this represent and what would Level 4 look like for this team?

This design team has never shipped an AI feature. Prescribe a 90-day plan to move them from Level 1 to Level 2 maturity. Be specific. What to build, what to measure, and what to stop doing.

Evaluate this AI product's approach to handling edge cases and failures. Score it on the maturity model. What does systematic (Level 3) failure design look like for this product specifically?