Stacks — Prompts

PACK 01 · USER RESEARCHERS

The Synthesis Stack

A 3-prompt chain that quickly extracts insights from raw user research data. Run them in sequence — summarise, theme, decide.

Summarise the Raw Transcript

You are a UX researcher. Here's a transcript of a user interview: [insert]. Summarise the key goals, frustrations, and behaviours mentioned by the user in bullet points.

→ A clear summary, fast

Find Themes Across Interviews

Here are summaries from 5 user interviews: [insert]. Identify recurring patterns and themes across these participants. Group findings under headings like 'Common Goals', 'Pain Points', and 'Unmet Needs'.

→ An affinity map without post-its

Extract Insight Statements

Based on the themes above, generate 3–5 concise insight statements that summarise user behaviour and motivation. Follow this format: 'Users [do something] because [reason].'

→ Insights, ready for ideation

From the pack User Researchers →

PACK 02 · PRODUCT DESIGNERS — 0→1

The 0→1 Stack

A 3-prompt chain that takes you from raw idea to a tested concept ready to build. Run them in sequence — frame, sketch, decide.

Frame the Problem Sharply

Act as a senior product designer. For this fuzzy idea: [insert], write a sharp 1-page problem brief. Cover: who it affects, when, why current solutions fail, what 'good' looks like. Resist the urge to design the solution.

→ A problem brief, not a solution

Generate Three Different Concepts

Generate 3 genuinely different concept directions for that problem. For each: core idea, how user experiences it, what's distinctive, what it sacrifices. Make them different mechanics, not three flavours of the same idea.

→ Three concepts to choose between

Decide What to Test First

Pick the concept with the most upside if right and the cheapest test. Justify the choice in plain language, then write the smallest test we could run this week to disprove it.

→ A decision and a test, ready to run

From the pack Product Designers — 0→1 →

PACK 02 · PRODUCT DESIGNERS — DELIVERY

The Iteration Stack

A 3-prompt chain that turns a falling metric into a designed iteration. Run them in sequence — diagnose, design, decide.

Decompose the Metric

Act as a growth analyst working with a designer. [Insert metric] is declining. Decompose it: break the metric into inputs, identify which input is moving, propose 5 plausible root causes (across content, flow, performance, comms, audience). Recommend the cheapest first investigation.

→ A diagnosis, not a hunch

Design the Iteration

For the most likely root cause, design 2 iteration options: a small change (cheap, fast) and a structural change (slower, bigger). For each: what changes, what risk, expected impact. Recommend which to ship first.

→ Two design options, ranked

Plan the Test and Decision Rule

Plan the A/B test for the recommended iteration. Specify: hypothesis, control, variant, success metric, guardrail metrics, sample size, decision rule. End with the rule for what we ship after the result.

→ A test plan with a decision baked in

From the pack Product Designers — Delivery →

PACK 03 · BUSINESS DESIGNERS

The Validation Stack

A 3-prompt chain that takes you from a raw business idea to a clear next move. Run them in sequence — assumption, test, decision.

Surface the Riskiest Assumption

Act as a lean strategist. For this business idea: [insert], list every assumption baked into it (desirability, feasibility, viability). Rank them by: a) how risky if wrong, b) how cheap to test. Identify the single riskiest, cheapest-to-test assumption — the one we should test first.

→ One sharp, testable assumption

Design the Cheapest Test

For that assumption, design the smallest, fastest, cheapest experiment that could disprove it. Specify: test type, what we measure, success threshold, failure threshold, time to run, cost to run. Output as a 1-page test card.

→ A test card, runnable this week

Convert Result Into Decision

Based on this test result: [insert], write a 1-page decision memo. Cover: what we tested, what we learned, what surprised us, what changed in our thinking, and the next decision: Persevere, Pivot, Pause, or Kill.

→ A decision, not just data

From the pack Business Designers →

PACK 04 · SERVICE DESIGNERS — DESIGNING

The Blueprint Stack

A 3-prompt chain that takes you from raw journey notes to a layered, useful service blueprint. Run them in sequence — spine, layers, fail points.

Build the Journey Skeleton

Act as a service designer. For this user goal and service: [insert], write the customer-facing journey skeleton. Use stages: Trigger, Find, Choose, Use, Pay, Get Help, Leave/Renew. For each stage, list the user actions and the channels involved. Keep it tight — this is the spine.

→ A clean journey spine

Layer the Blueprint

Take that journey skeleton and layer it. For each step, fill in: frontstage interactions, line of visibility, backstage actions, support processes, evidence, time, emotion. Present as a stacked blueprint.

→ A layered, practical blueprint

Surface the Fail Points

Across that blueprint, identify every fail point, bottleneck, and single point of failure. For each: which layer it sits in, what triggers it, what it does to the user, a 1-line mitigation. Rank by user impact.

→ A prioritised fail-point list

From the pack Service Designers — Designing →

PACK 04 · SERVICE DESIGNERS — RUNNING

The Recovery Stack

A 3-prompt chain that turns a service failure into a structural fix. Run them in sequence — frame, design, prevent.

Frame the Failure

Act as a service designer. For this service failure: [insert incident]. Reconstruct what happened from the user's perspective: what they expected, what they got, where it broke, how they found out, what they did next. Avoid blame — describe the experience.

→ A user-side failure narrative

Design the Recovery

Design the recovery protocol for users affected by this failure. Cover: how we tell them, what we say, what we do, the timeline, the goodwill gesture, who owns it, how we close the loop. Reference the 'easy to recover from mistakes' principle.

→ A recovery playbook

Prevent Recurrence

Identify the root cause behind the failure (frontstage, backstage, support, evidence, policy). Recommend the structural change that stops it recurring. Specify: what changes, who owns it, how we know it worked, the leading indicator we'd watch for the next failure.

→ A structural fix, owned

From the pack Service Designers — Running →

PACK 05 · FACILITATORS — GENERAL

The Workshop Rescue Stack

A 3-prompt chain for when a session is going off the rails mid-flow. Run them in sequence — diagnose, decide, script.

Diagnose the Drift

Act as a facilitator. The session is off track in [insert way]. Diagnose what's actually wrong: wrong people in the room, wrong activity for the goal, wrong energy, hidden conflict, scope confusion, or fatigue. Pick the most likely cause and explain why.

→ A clear diagnosis

Decide the Move

Based on that diagnosis, recommend the move: call a break, change activity, name the elephant, return to the outcome, restart, or end early. Justify the choice. List 1 risk of the move and how to mitigate.

→ A decision, fast

Script the Move

Write the literal words I should say in the next 60 seconds to execute that move. Tone: calm, direct, non-defensive. End with what I do as soon as I've said it.

→ Words ready to say

From the pack Facilitators — General →

PACK 05 · FACILITATORS — SPRINT

The Sprint Rescue Stack

A 3-prompt chain for when a sprint is going wrong mid-week. Run them in sequence — diagnose, decide, recover.

Diagnose Where It's Broken

Act as a senior sprint facilitator. The sprint is in trouble in [insert way — e.g. team disagrees, decider missing, scope unclear, prototype slipping]. Diagnose what's actually broken: the framing, the team, the schedule, the decider, the technique, the energy. Pick the most likely cause.

→ A diagnosis, not a panic

Decide the Move

Based on that diagnosis, recommend the move: pause and reframe, swap an activity, escalate to the decider, end early, redefine the test, or carry on. Justify. List 1 risk and how to mitigate.

→ A clear decision

Script the Recovery

Write the script for executing the recovery move. Cover: what I tell the team, how I frame it (not as failure), how I redirect the energy, what we do in the next 30 minutes. Tone: calm, direct, non-defensive.

→ A recovery, in motion

From the pack Facilitators — Sprint →

The Stacks