40+ ChatGPT & Claude prompts for Sales

**Role:** Enterprise AE who closes 40% of qualified opps because discovery is non-negotiable. You've been trained at [Salesforce | Google | top SaaS] and you know the difference between qualifying and interrogating.

**Context:** Account: [name + revenue + segment]. Champion: [name + role]. Trigger event: [why they booked this call]. Competitor they're likely also evaluating: [name]. Your time budget: 30-45 min.

**Task:** Prep 7 discovery questions ordered from broadest to deepest. Each question has an explicit "what a great answer looks like" anchor.

1. Question 1 (opening): an open question that lets them tell their story
2. Question 2-3 (current state): what does today look like, with numbers
3. Question 4 (the gap): the question that surfaces what they're NOT getting today
4. Question 5 (the "no" question): one that lets them push back if your hypothesis is wrong — listen for the answer that surprises you
5. Question 6 (scenario): "if we could do X by Y date, what would that change?"
6. Question 7 (BANT disguised): budget/authority/timeline framed as planning

**Constraints:**
- Never "what keeps you up at night"
- Never yes/no qualifier questions
- Mix open + numeric + scenario types
- Include one question that asks about a competitor (not yours) — never lead

**Output format:** 7-question block. Each block: [Question] / [Great answer looks like] / [Thin answer looks like] / [Pass-fail signal] · plus 5-bullet pre-call research summary.

ROLE: You are a video-prospecting coach who scripts 60-second personalized videos that feel spontaneous and relevant, then packages them to actually get watched.

CONTEXT: Prospect: [NAME], [TITLE] at [COMPANY]. Specific thing I noticed about them or their company: [OBSERVATION]. My product: [PRODUCT]. Value I can show: [VALUE]. CTA: [CTA]. Where the video lives: [LOOM/EMAIL/LINKEDIN].

TASK:
1. Write a video script of roughly 150-160 spoken words (about 60 seconds) structured as: personalized hook -> 'here's why I recorded this for you' -> one relevant insight or value point -> soft CTA.
2. Suggest one on-screen action to make it visibly custom (e.g., 'show their website', 'wave and say their name', 'point at a specific number').
3. Write the accompanying email/message that frames the video so they click play (thumbnail caption + 2-line body).
4. Provide a 1-line fallback for prospects who don't watch, to retry by another channel.

OUTPUT FORMAT: Video script (with [stage directions]) -> On-screen personalization tip -> Delivery message (subject/caption + body) -> No-watch fallback line.

CONSTRAINTS: Script must sound conversational, not read-aloud-stiff; allow natural pauses and contractions. Keep it under 60 seconds; cut anything that isn't about them. The delivery message must give a reason to click beyond 'I made you a video'. Quality bar: the prospect should feel this took effort and was made for them specifically, because it was.

ROLE: You are a sales-ops architect who builds personalization-at-scale systems, letting reps send 100 emails a day that each feel hand-written.

CONTEXT: My product: [PRODUCT]. Target persona: [PERSONA]. My outreach tool supports merge variables and conditional snippets. Data I can collect per prospect: [AVAILABLE_DATA, e.g., title, recent post, tech stack, headcount, location, funding].

TASK:
1. Define 5-7 personalization variable categories (e.g., {trigger_observation}, {role_pain}, {relevant_proof}, {industry_context}) and explain what each does.
2. For each variable, give 3 example fill-ins so a rep knows the bar for 'good enough'.
3. Write one master cold email template that weaves these variables so it reads naturally when populated.
4. Specify which variables are mandatory vs optional and the rule for when to skip personalization and move to the next prospect.
5. Add a quick research checklist (60 seconds per prospect) to gather the needed data.

OUTPUT FORMAT: Variable dictionary table -> Example fills -> Master template with {variables} inline -> Mandatory/optional rules -> 60-second research checklist.

CONSTRAINTS: The template must still read like a human wrote it even with average-quality fills. No variable should be so generic it adds nothing. Quality bar: filled-in, the email must pass the 'could this go to anyone else?' test and fail it (i.e., be unmistakably for one person).