A step-by-step way to work
Bring sound analysis, not raw data
SCR is for communicating a finding you've already validated — don't ask AI to find the story in unanalysed numbers.
Lay out S–C–R
Give AI your findings and ask it to arrange them as Situation (the context everyone agrees on), Complication (what changed/what's wrong), Resolution (your recommendation).
Cut to the story
Ask it to strip a dense draft to the few points that drive the decision.
Pressure-test the logic
'Does my resolution actually follow from my complication?' — AI is decent at catching logical gaps.
Own the recommendation
The resolution is a decision you're advocating; AI structures it, you stand behind it.
Real prompts, and what good output looks like
Turn these findings into an SCR story for a leadership review: trial-to-paid conversion is 12% (down from 18% last quarter); the drop is concentrated in self-serve signups; guided-onboarding signups still convert at 19%. I want to recommend adding guided onboarding to the self-serve flow.
Situation: We rely on trials to drive new revenue; last quarter trials converted to paid at 18%.
Complication: This quarter conversion fell to 12%, and the drop is entirely in self-serve signups — guided-onboarding signups still convert at 19%.
Resolution: Bring guided onboarding into the self-serve flow; the 7-point gap suggests recovering much of the lost conversion without adding headcount.
The logic holds IF the guided vs self-serve difference is causal, not just a difference in who chooses each path — worth a caveat or a test.
A prompt you can reuse
Fill in the highlighted parts and paste it into your AI tool of choice. Edit the output — it's a starting point, not a finished answer.
Help me turn validated findings into a Situation–Complication–Resolution data story. Don't invent data; do challenge my logic. My findings (already analysed): [paste the key numbers/findings] The decision I want to advocate: [your recommendation] Audience: [who's in the room] Please: 1. Structure it as S–C–R — Situation everyone agrees on, Complication that creates tension, Resolution as my recommendation. 2. Cut anything that doesn't advance the decision. 3. Pressure-test: does the Resolution actually follow from the Complication? Flag any leap (e.g. correlation treated as cause). 4. Suggest where I need a caveat or a confirming test before I present.
What AI gets wrong here
Narrating an unsupported conclusion
AI will build a clean story toward whatever resolution you give it, even if the data doesn't support it.
Persuasion over honesty
A slick SCR can make a weak case sound strong.
Losing the caveat in the polish
Tightening the story can quietly drop the 'this is correlational' honesty.
AI structures a data story well and — valuably — will catch logical leaps in your own argument if you ask it to. But it can't tell you what the data means or which decision is right; it will narrate whatever conclusion you hand it. The interpretation and the recommendation are yours, and the analysis behind them must be sound before you let AI make it sing.