AI prompts for CSMs

Good prompts save hours of prep. These templates cover the most common CSM tasks. Replace the bracketed parts, and never paste confidential customer data into a tool your company has not approved.

Quick answer — Which AI tool is best for customer success managers? Several general-purpose assistants work well for drafting and research. Use the tools your company approves, and prioritize ones that meet your data and security requirements.

Rules before you start

  • Check your company's AI policy before putting customer data into any tool.
  • Give context: role, audience, goal, tone, and constraints.
  • Ask for a draft, not a final. You own the result.
  • Verify facts and numbers. Models can state wrong things confidently.

Account prep and research

Prompt: You are a customer success manager preparing for a call with [customer name], a [industry] company using our [product]. Here are my notes and usage summary: [paste approved notes]. Summarize the account in 5 bullets, list likely risks and opportunities, and suggest 5 discovery questions.

Prompt: Based on this company description [paste public info], suggest likely business priorities for this year and how our product could connect to them. Flag anything uncertain.

QBR and EBR drafting

Prompt: Draft an outline for a [QBR/EBR] for a customer whose goals are [goals]. Usage trend: [trend]. Wins: [wins]. Open issues: [issues]. Structure it as: recap, business impact, risks, roadmap alignment, next steps. Keep it executive-friendly. See QBRs and EBRs.

Emails and follow-ups

Prompt: Write a concise, warm renewal outreach email to [name, title], 90 days before renewal. Reference these outcomes: [outcomes]. Ask for a 30-minute call. Avoid hype and discount language.

Prompt: Turn these meeting notes into a recap email with decisions, owners, and dates: [notes]. Keep it under 150 words.

Risk and health summaries

Prompt: Here are the last 6 months of support tickets, usage notes, and stakeholder changes for [customer]: [paste approved data]. Identify churn risk signals, rank them, and suggest a 30-day save plan. See the at-risk playbook.

Success plans and playbooks

Prompt: Create a customer success plan for a new [segment] customer whose goal is [goal] and success criteria are [criteria]. Include milestones for the first 90 days, owners, and risks. See the success plan template.

More prompts for common situations

Prompt: I am a CSM. A key stakeholder at [customer] has gone quiet for 6 weeks. Here is the account context: [paste approved context]. Suggest 3 outreach angles, each with a short email draft, and rank them by likelihood to get a reply.

Prompt: Summarize this call transcript into: 1) customer goals mentioned, 2) risks or concerns, 3) expansion signals, 4) commitments I made, 5) recommended next steps. Transcript: [paste approved transcript].

Prompt: Help me prepare for a renewal negotiation. The customer is asking for a 20% discount. Context: [value delivered, usage, relationship]. List the likely objections, a response for each, and 3 options I can offer that are not a straight discount.

Prompt: Turn these notes about a product gap into a clear, concise feedback entry for the product team: the customer problem, who is affected, business impact, and examples. Notes: [paste].

Prompt patterns that improve results

PatternExample
Role + audience"You are a CSM writing to a skeptical VP of Operations."
Constraints"Under 120 words, no jargon, no discount language."
Structure"Format as: situation, risks, plan, asks."
Examples"Here is an email in the tone I want: [paste]."
Critique step"List what is weak or missing in your draft, then rewrite it."
Verification"Flag every claim that I should verify before sending."

A review checklist before you send anything

  1. Are all names, numbers, and dates correct?
  2. Would the customer recognize this as written by you, in your voice?
  3. Does it contain any confidential information that should not leave your company?
  4. Does it make any commitment you cannot keep?
  5. Is the next step clear and small?

See also AI for account research and sentiment analysis.

Getting better output

  • Provide a good example of the format you want.
  • Ask the model to ask clarifying questions first.
  • Iterate: "make it shorter", "more direct", "remove jargon".
  • Keep a personal prompt library for repeat tasks.

For where this is heading, see how AI is changing CS and AI meeting summaries.

Frequently asked questions

Which AI tool is best for customer success managers?
Several general-purpose assistants work well for drafting and research. Use the tools your company approves, and prioritize ones that meet your data and security requirements.
Is it safe to put customer data into an AI tool?
Only if your company has approved the tool and its data handling. Check policy first and avoid pasting sensitive details into unapproved tools.
Can AI write my QBR?
It can draft structure and narrative from the data you provide, but you must verify facts and add judgment about what matters to this customer.
How do I write better AI prompts?
Provide context, audience, goal, format, and constraints, and iterate on the output.
What makes a good AI prompt for a CSM?
Give the model a role, audience, goal, the approved context, constraints on length and tone, and ask it to flag claims to verify. Then review and edit the result.
Can I use AI to write customer emails?
Yes, as a first draft, if your company allows the tool. Always review for accuracy, tone, and confidentiality before sending.

See what CSMs really earn

Compare your pay against real CSM submissions — free, anonymous, takes 2 minutes.

B
Ben Hancock
Customer Success Manager · built this database to bring transparency to CS compensation