Le générateur de questions d'entretien pour études de cas : transformer une réussite client vague en récit publiable
Pourquoi ce prompt est important
A B2B case study that only says a customer is “very happy” and cites one vague metric gets skimmed and forgotten; one with a specific incident, a real number, and a human quote gets forwarded internally during a prospect's own buying decision. A badly run 30-minute interview is unrecoverable — there's rarely a second chance to ask a customer executive for their time, so the cost of generic questions isn't a worse case study, it's often no usable case study at all.
À quoi nous l'utilisons
A product marketer at a mid-size SaaS company has a 30-minute call scheduled with a VP of Operations at a customer whose onboarding time reportedly dropped significantly after adopting their tool, and needs to walk into that call with a question list that won't waste the limited time on vague answers.
Prompt
Role: You are a B2B content strategist who has run hundreds of customer case study interviews and knows that generic questions produce generic case studies nobody reads past the headline. Context: My company is [YOUR COMPANY/PRODUCT]. The customer I'm interviewing is [CUSTOMER NAME, INDUSTRY, COMPANY SIZE]. Before working with us, their situation was: [DESCRIBE THE PROBLEM OR PAIN POINT AS YOU UNDERSTAND IT]. The result we believe they got is: [KNOWN METRIC OR OUTCOME, e.g., "cut onboarding time by 40%" — even if approximate]. The interview is scheduled for [LENGTH OF INTERVIEW, e.g., 30 minutes] with [WHO, e.g., "the VP of Operations who championed the purchase"]. Task: Build a structured set of interview questions that will extract a genuinely compelling, specific narrative — not vague praise. The questions should surface concrete numbers, emotional stakes, and quotable moments, organized so I can run the interview in order without losing the thread. Constraints: - No leading questions that invite a one-word "yes, it was great" answer. Every question should require the interviewee to describe something specific. - Include at least 2 questions designed specifically to surface a quotable, human sentence — not a metric, a feeling or moment (e.g., what did their boss say, what did they expect to happen vs. what actually happened). - Include at least 2 questions that probe for the cost of NOT solving the problem — what was it costing them before, in time, money, or stress — since this is usually what makes a case study persuasive to a prospect in the same situation. - Do not ask about product features directly. Ask about their problem and decisions; let the product come up naturally in their answers. - Flag which questions are "if time allows" versus "must ask" given my stated interview length, so I don't run out of time on the essentials. Output format: 1. **Before/Problem Questions** (must-ask) — 3-4 questions establishing the pain, its cost, and why they finally decided to act. 2. **Decision Questions** (must-ask) — 2-3 questions about evaluation and why they chose this path over alternatives (including doing nothing). 3. **Implementation Questions** (if time allows) — 2 questions about what the rollout/adoption was actually like, including any friction. 4. **Results Questions** (must-ask) — 2-3 questions pushing for specific numbers, timeframes, and what changed as a direct result. 5. **Quote-Mining Questions** (must-ask) — 2 questions aimed purely at getting one vivid, human sentence I can pull as a pull-quote. 6. **Suggested Case Study Structure** — a one-paragraph outline of how to sequence the eventual write-up (which answers become which section), so I know what I'm listening for during the interview itself.
Résultat
**Before/Problem Questions (must-ask)** 1. Walk me through a specific week before you started using [product] — what did your team's process actually look like, step by step? 2. What was this problem costing you — not in the abstract, but in hours, missed deadlines, or money you can point to? 3. Was there a specific moment or incident that made leadership finally say "we have to fix this now"? 4. What had you already tried before this, and why didn't it stick?
**Decision Questions (must-ask)** 5. Who else was in the room when you evaluated options, and what was the internal disagreement, if any? 6. What almost made you choose a different path — including doing nothing and living with the problem? 7. What was the one thing that tipped the decision in our favor, even if it seems small?
**Implementation Questions (if time allows)** 8. What was harder than you expected in the first two weeks? 9. Was there a point where you doubted this was going to work? What changed your mind?
**Results Questions (must-ask)** 10. Give me a specific number — time saved, cost avoided, output increased — and the timeframe it happened over. 11. What changed for your team day-to-day that you didn't expect going in? 12. If you had to put a dollar figure on what this is worth to you now, what would it be and how did you arrive at it?
**Quote-Mining Questions (must-ask)** 13. If you had to describe this to a peer at another company in one sentence, what would you say? 14. What did your boss or team say when they saw the results — do you remember their actual words?
**Suggested Case Study Structure** Open the write-up with the answer to question 3 (the specific incident) as a cold open — it's concrete and creates stakes immediately. Follow with questions 1-2 and 8 compressed into a "the situation" section that quantifies the pain without yet naming the solution. Questions 5-7 become "why they chose us," kept short — readers care less about the evaluation process than the outcome. Questions 10-12 are the core "results" section and should lead with the hardest number you got. Close with the answer to question 13 or 14 as a pull-quote, since it's the most human, shareable line in the entire interview.
La plupart des études de cas B2B échouent avant même d'être rédigées, dès l'entretien. Un responsable marketing obtient 30 minutes avec un dirigeant client très occupé, demande « pouvez-vous me parler de votre expérience avec nous ? » et reçoit exactement ce que cette question mérite : une réponse vague et polie qui ne dit rien de concret. Ce prompt existe parce qu'une étude de cas se gagne ou se perd pendant l'entretien, pas pendant la rédaction.
Pourquoi la structure place les questions difficiles en premier
Les questions sur la situation initiale/le problème et les questions sur les résultats sont marquées comme obligatoires, tandis que les questions sur la mise en œuvre sont explicitement signalées comme facultatives. Ce n'est pas arbitraire : dans un entretien limité dans le temps, les sections qui produisent le véritable récit (la douleur, le chiffre) comptent plus que l'acte intermédiaire (le déroulement du déploiement). La plupart des gens interrogent instinctivement d'abord sur la mise en œuvre, car cela semble être la question chronologiquement évidente, puis manquent de temps pour les deux sections qui rendent réellement une étude de cas convaincante.
Pourquoi le prompt interdit les questions sur les fonctionnalités du produit
Demander « qu'appréciez-vous dans la fonctionnalité X ? » produit des réponses qui ressemblent à une liste de fonctionnalités digne d'un texte marketing — parce que c'en est essentiellement une. Interroger plutôt sur le problème et les décisions du client fait que le produit est mentionné naturellement, avec les propres mots du client, ce qui est à la fois plus crédible pour le lecteur et généralement plus précis que tout ce qu'un responsable marketing aurait pu écrire lui-même.
Pourquoi deux questions sont réservées uniquement à la recherche de citations
La plupart des questions de ce prompt sont conçues pour extraire des informations : chiffres, délais, critères de décision. La section questions de recherche de citations est différente — elle vise à obtenir une seule phrase. « Qu'a dit votre patron en voyant les résultats ? » n'est pas une question informative ; c'est une demande portant sur un moment humain précis qui devient la citation phare que les rédacteurs utiliseront réellement. Séparer cette section en une catégorie à part permet qu'elle ne soit pas éclipsée par les questions plus analytiques posées plus tôt dans l'entretien.
Comment utiliser le résultat
Ne lisez pas les questions mot pour mot comme un script : utilisez-les comme une liste de vérification pour vous assurer d'avoir couvert l'essentiel, et laissez la conversation suivre son cours naturellement entre elles. La structure d'étude de cas suggérée à la fin est pensée pour être lue avant l'entretien, pas après : savoir à l'avance que la réponse à la question 3 constituera votre accroche signifie que, pendant l'entretien, vous êtes à l'écoute d'un incident précis et exploitable, plutôt que d'essayer d'en reconstruire un après coup à partir de notes.