Resources

AI workflows

AI prompts for founder-led sales

Copy-paste AI prompts for founder-led sales: thread summaries, follow-up drafts, objection handling, CRM notes, and weekly pipeline review.

Jun 29, 202610 min read

Prompting principles for sales work

Founder-led sales prompts should be specific, grounded in the thread, and easy to verify. The goal is not to produce impressive copy. The goal is to reduce the effort of doing careful follow-up work.

Treat AI output as a draft. Check names, dates, numbers, commitments, and anything that sounds like a promise before sending or saving it.

  • Ask for structured output so you can turn it into CRM notes or tasks.
  • Limit drafts by length so they do not become generic essays.
  • Require one concrete detail from the thread.
  • Ask the model to call out uncertainty instead of guessing.

Prompts for understanding a sales thread

Fast thread summary

Summarize this thread for founder-led sales. Return: buyer goal, current status, key people, objections or risks, promised follow-ups, and the next action.

Decision context

Based only on this thread, what decision is the buyer trying to make? List what they know, what they still need, and the best next question I should ask.

Stalled thread diagnosis

This sales thread has gone quiet. Diagnose the likely reason using only thread evidence. Then suggest one respectful follow-up and one reason not to follow up yet if applicable.

Prompts for follow-up drafts

Short follow-up

Draft a follow-up under 100 words. Use a calm founder tone, reference one specific detail from the thread, ask one clear question, and avoid pressure.

Proposal follow-up

Draft a proposal follow-up. Ask whether scope, timing, or next steps need clarification. Do not add new claims. Keep it concise and easy to reply to.

Close-the-loop email

Draft a polite close-the-loop email. Make it clear I am closing this on my side for now, and they can reply later if priorities change.

Prompts for CRM notes and tasks

CRM note

Create a CRM note from this thread with: context in one sentence, stage recommendation, next step, owner, due date if mentioned, and risks. Mark unknowns as unknown.

Task extraction

Extract action items from this thread. For each item, include owner, due date, dependency, and the next message I should send if I own it.

Stage recommendation

Suggest the best pipeline stage from: New, Qualified, Meeting, Proposal, Decision, Closed, Dormant. Explain the reason in one sentence and flag ambiguity.

Prompts for objections and positioning

Objection map

List objections or concerns in this thread. For each one, identify whether it is about fit, timing, budget, trust, urgency, or implementation. Suggest a response angle without inventing facts.

Clarifying question

Write 3 possible clarifying questions I can ask. Each question should be short, non-defensive, and useful for understanding the buyer concern.

Recap before decision

Draft a neutral recap before a decision. Include what we discussed, open questions, and the next decision point. Do not oversell.

Weekly review prompt

Once a week, use AI to prepare the review, then make the decisions yourself. The prompt should surface stuck work rather than pretend to close it for you.

Weekly pipeline review

Review these pipeline notes and group items into: needs my reply, waiting on them, proposal pending, unclear next step, and should be marked dormant. For each item, suggest one next task.

Before acting on the output, verify it against the actual Gmail threads and your current priorities.

Want this workflow inside Gmail?

Donna CRM runs inside Gmail as a Chrome extension. Use these workflows with real contact context, pipeline stages, and follow-up tasks - without leaving your inbox.

FAQs

Common questions about using AI prompts for founder-led sales.

Can I paste full sales threads into AI prompts?
Use the minimum context needed and follow your privacy obligations. If you use a product connected to Gmail, review its privacy terms and controls before relying on it for sensitive work.
How do I keep AI follow-ups from sounding generic?
Require one specific thread detail, keep the draft short, and edit the final message yourself. Generic output usually comes from vague prompts and too much length.
Should AI decide my pipeline stage?
It can suggest a stage, but you should confirm it. Stage changes should reflect real buyer commitment or process movement, not weak signals.

Related reading

Keep exploring: these pages go deeper on the feature set and the core Gmail CRM workflow.