Le Mineur de Témoignages Clients : Transformez des Transcriptions d’Entretiens Brutes en une Banque de Témoignages Catégorisés
Why this prompt matters
Most teams either skip customer interviews entirely because mining them for usable quotes takes hours, or they use one obvious quote from the interview and ignore four other equally strong ones buried deeper in the transcript. A systematic extraction pass turns 45 minutes of raw audio-to-text into a reusable quote bank that can supply a landing page, three sales one-pagers, and a quarter of social posts from a single round of interviews — instead of running a new interview every time a different team needs a quote.
What we use it for
You just finished five 30-minute customer interviews for a case study project. You have five raw transcripts, roughly 45 minutes of usable content buried in filler words, tangents, and repeated phrasing, and marketing needs three strong quotes for a landing page redesign by Friday.
Prompt
Act as a content strategist specializing in customer proof and testimonial curation for [YOUR COMPANY/PRODUCT NAME].
Context:
- Product/service being discussed: [BRIEF DESCRIPTION OF WHAT YOU SELL]
- Common objections your sales/marketing team currently struggles to address: [LIST 3-5, e.g., "too expensive for small teams", "hard to migrate existing data", "not sure it works for our industry"]
- Customer segment(s) in the transcripts: [E.G., "Enterprise IT buyers" or "Solo freelancers" — note per transcript if mixed]
Raw transcripts:
[PASTE ONE OR MORE RAW INTERVIEW TRANSCRIPTS, SUPPORT CALL NOTES, OR REVIEW TEXT HERE — filler words and interviewer questions are fine, leave them in]
Task:
1. Extract every genuinely usable testimonial-quality quote from the transcript(s) — a sentence or short passage that clearly and specifically praises, defends, or explains a benefit of the product in the customer's own voice.
2. For each extracted quote:
- Lightly clean it for grammar and remove filler words ("um," "like," false starts) WITHOUT changing the customer's actual meaning, claims, or tone
- Tag it with the SPECIFIC objection or use case it addresses (not a generic category — e.g., "addresses price-sensitivity for small teams" not just "pricing")
- Note the customer segment it came from, if provided
- Rate its strength: Strong (specific, vivid, quotable) / Moderate (useful but generic) / Weak (true but not compelling enough to use standalone)
3. Flag any quotes that make a specific measurable claim (a number, a percentage, a time saved) — these are the highest-value quotes and should be surfaced first.
4. If NONE of the provided objections are addressed by any quote in the transcript, say so explicitly rather than forcing a weak match.
Constraints:
- Never alter, exaggerate, or combine quotes from different parts of the transcript into a single fabricated quote — every quote must be traceable to one continuous passage in the source text.
- Do not invent a customer's job title, company, or any attribution detail not present in the transcript.
- Skip quotes that are too vague to be useful ("it's great," "I like it") even if grammatically clean.
Output format:
- Group quotes under headers by the objection/use-case tag they address
- Within each group, order quotes by strength (Strong first)
- Format each entry: cleaned quote (in quotation marks) — Strength rating — Segment (if known) — one-line note on where it could be used (landing page hero, sales one-pager, social post, etc.)
- A closing section titled "Objections Not Yet Addressed" listing any objections from your list with no supporting quote foundResult
Addresses: Onboarding-speed skepticism ("too complex to set up")
1. "I was honestly expecting a two-week onboarding slog like our last tool, and we were running real reports by day three. That alone justified switching." — Strength: Strong — Segment: Mid-market operations lead — Best for: landing page hero quote, sales one-pager top section (specific number, direct comparison to switching cost)
2. "Setup wasn't instant, but it also wasn't the nightmare I'd built up in my head from other implementations." — Strength: Moderate — Segment: Mid-market operations lead — Best for: FAQ section addressing setup concerns directly, less punchy for a hero placement
Addresses: Price-sensitivity for small teams
3. "We're a six-person team and I genuinely wasn't sure we could justify an enterprise-tier tool. Turns out the core features we actually use are all in the starter plan — we're paying less than our old spreadsheet-plus-three-plugins setup cost us." — Strength: Strong — Segment: Small business owner — Best for: pricing page testimonial, small-business-targeted social post (specific comparison, addresses the exact fear directly)
Addresses: Migration/data-import concerns
4. "Moving four years of customer records over was the part I was dreading most, and it took about twenty minutes with the import tool. I kept waiting for something to break." — Strength: Strong — Segment: Enterprise IT buyer — Best for: sales one-pager for technically cautious buyers, case study opening line (specific time claim, addresses a named fear directly)
Addresses: Industry-fit doubt ("not sure it works for our industry")
5. "Most of the case studies I'd seen were from SaaS companies, and we're a physical logistics operation, so I wasn't sure this would map to how we actually work. It took maybe a week of light customization before it felt built for us specifically." — Strength: Moderate — Segment: Logistics/operations — Best for: industry-specific landing page variant, not strong enough alone for the main homepage (true and specific, but less punchy/quotable than the top quotes)
Measurable claims flagged for priority use
- Quote #1 — implied setup time comparison (2 weeks expected vs. 3 days actual)
- Quote #3 — cost comparison (cheaper than prior spreadsheet-plus-plugins stack)
- Quote #4 — specific time claim (20 minutes for a 4-year data migration)
Objections Not Yet Addressed
No quote in the provided transcripts addresses "lack of customer support responsiveness" — this objection from your list has no supporting evidence in this batch. Recommend flagging this for the next round of customer interviews specifically, or pulling from support ticket resolution data instead.
Les entretiens clients sont coûteux à réaliser – planifier, mener et transcrire prend un temps considérable – et la majeure partie de cette valeur est gaspillée car la transcription obtenue fait 30 à 45 minutes de bavardage décousu avec peut-être cinq phrases vraiment citables disséminées dedans. Trouver ces cinq phrases en relisant l’intégralité de la transcription est si fastidieux que la plupart des équipes soit ignorent l’exercice, soit se contentent de la citation dont elles se souviennent de l’appel.
Ce Que Fait Ce Prompt
Ce prompt agit comme un stratège de contenu qui effectue une extraction systématique de citations. Vous collez une ou plusieurs transcriptions brutes, et il extrait des citations de témoignages réellement utilisables, chacune nettoyée grammaticalement sans modifier les mots réels du client ni leur sens, et étiquetée en fonction de l’objection ou du cas d’usage spécifique qu’elle aborde – scepticisme sur le prix, doute sur la facilité d’installation, comparaison avec un concurrent, etc. Au lieu d’une citation générique, vous obtenez une banque catégorisée que vous pouvez associer à la page ou à la conversation de vente qui en a besoin.
Pourquoi la Catégorisation est Importante
Un témoignage qui dit « cela nous a fait gagner du temps » est bien moins utile qu’un témoignage étiqueté spécifiquement comme « aborde le scepticisme sur la rapidité d’intégration » lorsque vous créez une page destinée à des acheteurs inquiets du temps de mise en œuvre. Le prompt impose cette spécificité en exigeant une étiquette explicite d’objection ou de cas d’usage pour chaque citation, ce qui transforme un tas de phrases agréables à entendre en citations que vous pouvez réellement déployer face à une hésitation particulière de l’acheteur.
Comment l’Adapter
Collez les transcriptions telles quelles – le prompt est conçu pour gérer les mots de remplissage, les faux départs et les questions de l’intervieweur mélangés dans le même bloc de texte. Si vous avez des entretiens provenant de plusieurs segments de clients (par exemple, entreprise vs petite entreprise), notez le segment pour chaque transcription afin que la sortie puisse indiquer quelles citations fonctionnent le mieux pour quel public.