Write a case study from a raw customer interview transcript
Turns a raw, rambling customer interview into a publish-ready case study — result-led headline, snapshot metrics box, four-part story, and verbatim pull quotes — while refusing to invent quotes or numbers. If the transcript is missing a strong metric, it tells you exactly what to ask the customer instead of papering over the gap.
You are a B2B case study writer whose work gets forwarded inside buying committees. You know the two ways case studies die: they read like press releases ('Company X leveraged our innovative platform...'), or they bury the one number that matters under 800 words of setup. Your rules: the customer is the hero, their words beat your paraphrase, and every claim traces back to the transcript.
I'll give you a raw customer interview transcript. Produce:
1. A headline with the strongest specific result in it (number if one exists).
2. A snapshot box: company, industry, size, use case, and 2-3 headline metrics.
3. The story in four sections — Before (the pain, in their words), Why they chose us, How it rolled out, Results — 500-700 words total.
4. Three pull quotes, verbatim from the transcript, chosen for specificity over flattery.
5. A closing CTA line I can adapt.
Hard rules: never invent or embellish a quote — verbatim only, light cleanup of filler words allowed. Never extrapolate a number the customer didn't say. If the transcript lacks a strong metric, flag it and list the exact follow-up questions I should ask the customer to get one. Write at an 8th-grade reading level. No 'delighted to partner with'.
Before you write anything, interview me. Ask me these questions ONE AT A TIME, waiting for my answer each time:
1. Paste the full interview transcript.
2. What does your product do, in one or two sentences, for context?
3. Who is the target reader of this case study — same industry as the customer, same role, or a specific deal you're trying to influence?
4. Any constraints — can you name the customer, are any numbers off-limits, is there a word count target?
Once you have my answers, write the case study. If the transcript contradicts itself anywhere, ask me which version is accurate rather than picking one silently.How to use it
- 1
Copy the prompt into Claude, ChatGPT, or any LLM with a large context window.
- 2
Export the interview transcript from Zoom, Gong, or Fathom and paste it in full — don't pre-trim it.
- 3
Answer the reader-targeting question honestly; a case study aimed at one live deal reads differently than a generic one.
- 4
Check every pull quote against the transcript before publishing — verbatim means verbatim.
- 5
Send the draft to the customer for approval, then rerun the prompt asking for a one-paragraph version for proposals.
Best practices
Ask customers 'what were you doing before?' and 'what would you tell someone considering us?' in the interview itself — those two answers write most of the case study.
If the model flags a missing metric, actually send the customer the follow-up questions it lists; a single number can double the asset's usefulness.
Rerun with question 3 changed to generate industry-specific edits of the same story for different sequences.
Keep the snapshot box metrics to three or fewer — a wall of stats reads less credible than two sharp ones.
Example: what this looks like in practice
A marketer at a 45-person manufacturing-software company pastes a 40-minute customer interview transcript from Fathom. The customer, a plant manager, mentions offhand that unplanned downtime dropped 'from about nine hours a month to two, maybe three'. The model leads with 'Cutting unplanned downtime from 9 hours a month to 3', builds the snapshot box, and pulls three quotes including the skeptical one — 'honestly I assumed it was another dashboard we'd ignore' — which becomes the Before section's spine. It flags that no dollar figure exists and suggests two follow-up questions. The customer replies with '$14K per downtime-hour', and the final study anchors both numbers. Sales starts attaching it to every mid-funnel manufacturing deal.
Best fit
This prompt is one gear in a bigger machine. We orchestrate 20+ tools into outbound systems our clients own — and guarantee the results.
Apply for a Pilot Spot → →Frequently asked questions
Yes, and it's one of the highest-leverage uses of AI in marketing — the raw material already exists, the work is structure and selection. The key is constraining it: this prompt requires verbatim quotes only, bans extrapolated numbers, and forces the model to flag gaps rather than fill them with plausible-sounding fiction. You verify quotes against the transcript before publishing.
More sales copywriting & assets prompts
Generate an outcome-first value proposition buyers can repeat
Builds a complete value proposition system — primary line, three angle variations, a disqualifier, and channel-ready versions — instead of one throwaway tagline. The interview forces buyer, outcome, mechanism, and proof out of your head first, so the output is positioning you can defend, not word salad you'll rewrite next quarter.
Align your cold email and landing page so the click doesn't break the story
Audits your cold email and landing page side by side, finds where the story breaks between click and page, and fixes the cheaper side with minimum edits. The 'message spine' it outputs becomes a five-line contract both assets obey, so the next person who edits either one doesn't silently reintroduce the mismatch.
Build a sales deck narrative arc: problem, shift, proof, ask
Structures your pitch into a 10-12 slide narrative arc — problem, shift, proof, ask — with full-sentence headlines that carry the argument on their own, evidence notes, and talk-track lines per slide. It replaces the product-tour deck with a story that ends in a decision, plus a 60-second fallback when the meeting gets compressed.