Create a copy-paste AI prompt pack for a salesperson selling [PRODUCT OR SERVICE] to [TARGET BUYER]. My sales motion and guardrails: [SALES CONTEXT]. Product facts and approved proof: [PRODUCT FACTS AND PROOF]. Current account or opportunity details: [ACCOUNT DETAILS]. Desired outcome: [DESIRED OUTCOME]. Return exactly eight standalone prompts, each 110–150 words, written in first person so I can paste one into ChatGPT, Claude, or Gemini. Each prompt must include its own square-bracket inputs and tell the AI exactly what to return. Cover these jobs in this order: 1. Account research brief. 2. Stakeholder map and buying-committee hypotheses. 3. Discovery-call plan and questions. 4. Personalized first outreach email. 5. Call recap and follow-up email. 6. Objection-preparation sheet. 7. Mutual action plan for a live opportunity. 8. Deal-risk review and next-best action. Build in the discipline a strong salesperson uses: separate verified account facts from assumptions; tie a proposed outcome to a stated business problem; avoid invented customer logos, ROI, integrations, security claims, pricing, timelines, or executive quotes; and do not manufacture urgency. For outreach, require a concrete reason for contacting that person and one clear call to action. For discovery, include problem, impact, current process, stakeholders, decision criteria, procurement, timeline, and next-step questions without turning them into a rigid script. After the eight prompts, add a short “claim-control checklist” with six checks I should make before sending any customer-facing copy. Before answering, verify that every prompt is usable without access to private CRM data and that each prompt contains its own needed placeholders. Ask up to 3 clarifying questions only if a required input is missing.
Fill in
| Placeholder | What to enter | Example |
|---|---|---|
| [PRODUCT OR SERVICE] | Enter what you sell, the use case it addresses, and its delivery model. | Cloud software that automates vendor-security questionnaire workflows for mid-market companies. |
| [TARGET BUYER] | Enter the buyer roles, company profile, and problem areas you sell to. | VPs of security, GRC directors, and sales operations leaders at B2B SaaS companies with 300–2,000 employees. |
| [SALES CONTEXT] | Enter sales cycle, typical deal size, territory or segment, approved messaging rules, and any CRM or privacy restrictions. | Outbound and inbound enterprise motion; 60–120 day cycle; $25k–$75k annual contracts; no claims of guaranteed audit outcomes. |
| [PRODUCT FACTS AND PROOF] | Enter approved capabilities, limitations, integrations, pricing boundaries, case studies, and proof points. | Imports approved answer libraries, routes reviews, and tracks response status; integrates with Google Drive and Slack; SOC 2 Type II; pricing only after discovery. |
| [ACCOUNT DETAILS] | Enter the known account facts, stakeholders, opportunity stage, prior interactions, and stated pains. | Northstar HR, 650 employees; security director downloaded a questionnaire template; SDR noted they answer about 40 questionnaires per quarter; no meeting booked. |
| [DESIRED OUTCOME] | Enter the immediate goal, such as booking discovery, preparing a renewal meeting, or rescuing a stalled deal. | Book a 25-minute discovery call with the security director. |
How to use
- Fill in approved product facts and real account details before generating the pack; leave unknown fields as unknown.
- Start with the account-research and discovery prompts, then use the call-recap prompt immediately after a conversation while notes are fresh.
- Check every customer-facing line against your approved proof, especially outcomes, integrations, security, pricing, and named customers.
- Follow up with: “Adapt prompts 4 and 6 for a renewal stakeholder who is concerned about adoption, using only these notes: [NOTES].”
Variations
Discovery prep
Use before a first or second discovery conversation.
Prepare a discovery-call plan for [ACCOUNT] and [CONTACT]. Known facts: [ACCOUNT FACTS]. What we sell: [PRODUCT FACTS]. Meeting goal: [MEETING GOAL]. Produce a 30-minute agenda, 12 sequenced questions across problem, impact, current workflow, stakeholders, decision process, risk, and timing, plus listening cues and two possible next steps. Do not assume budget, authority, or a pain point that is not evidenced. Flag claims I need to validate in the call.
Follow-up email
Use after a customer conversation when you have specific notes and an agreed next step.
Draft a post-call sales follow-up from these notes: [CALL NOTES]. Customer role and company: [CONTACT DETAILS]. Approved product facts: [APPROVED FACTS]. Proposed next step: [NEXT STEP]. Write a 140-word email with a factual recap, two customer-stated priorities, relevant capability only where supported, owner and date for the next step, and one clear CTA. Do not add a business case, pricing, timeline, or commitment absent from the notes. List any ambiguous statement for my review.
Deal inspection
Use for a qualified opportunity that has slowed or lost momentum.
Run a deal inspection for [OPPORTUNITY]. CRM facts: [CRM FACTS]. Conversation notes: [CONVERSATION NOTES]. Sales process: [SALES PROCESS]. Produce a table for problem, impact, champion, economic buyer, technical fit, decision criteria, procurement, competition, timeline, and next commitment. Mark each as verified, hypothesis, or missing. Then recommend three evidence-gathering actions and a respectful close-lost path if appropriate. Do not invent urgency or label a contact a champion without evidence.
Tips
- Use real customer language from notes in follow-ups, but do not turn an exploratory comment into a confirmed requirement.
- A useful stakeholder map labels assumptions; it does not pretend that a job title proves authority, influence, or opposition.
- After each call, capture the customer’s problem, impact, decision process, and next commitment separately. “Good conversation” is not a sales next step.
- Keep an approved proof library for metrics, integrations, security statements, and case studies so generated copy cannot accidentally make unsupported claims.
FAQ
Can AI write prospecting emails for salespeople?
Yes, if you supply real account research and approved product claims. It should produce a relevant reason to contact the prospect, not invent personalization.
What sales tasks should I avoid delegating to AI?
Do not let it invent pricing, contract language, security assurances, customer references, or what a buyer said. Verify CRM updates and customer-facing commitments yourself.
How do I prevent generic discovery questions?
Provide the account’s situation and desired meeting outcome, then ask for questions that test specific hypotheses rather than a universal discovery checklist.