Call notes to CRM update prompt: free ChatGPT template
Updated · By Robert Breen
Use this straight after a sales call, when your notes are half sentences and the CRM entry is the last thing you want to write. The prompt sorts the notes into the fields a CRM record needs: a two-line summary, what the buyer needs, objections, who else is involved, and one next step with a date. It leaves the deal stage and amount to you, because those are judgment calls and a guessed number in the pipeline is worse than a blank.
The prompt
Copy it into ChatGPT (or Claude, Gemini or Copilot) and replace every [BLANK] with your own details. It uses the four parts from Prompt Writing 101: Role, Context, Task and Format.
Role: You are a sales operations assistant who keeps CRM records short and accurate. Context: I just finished a call with [CONTACT] at [COMPANY]. We sell [WHAT WE SELL]. These are my raw notes, which may be incomplete: [NOTES]. Task: Turn the notes into a CRM update with these fields: summary (two sentences), what they need, objections or concerns, other people involved, next step, next step date, and open questions. Use only what the notes say. If a field has no information, write "not discussed". Do not set a deal stage, a close date or an amount. Flag anything in the notes that sounds like a promise we made. Format: Field name, a colon, then the entry, one field per line, ready to paste into the CRM. No bullets inside fields. After the fields, one line starting "Promises to check:".
Fill in the blanks
[CONTACT]- The person's name and role.
[COMPANY]- Their company.
[WHAT WE SELL]- The product or service in a few words.
[NOTES]- Your raw notes from the call, typos and all. Don't clean them first.
Example, filled in
A made-up example from the free lesson Build an n8n AI Agent That Writes Deal Follow-Ups. Northwind Scheduling Software is the made-up software company from the sales lessons. A made-up account executive has just finished a discovery call and needs to log it.
Role: You are a sales operations assistant who keeps CRM records short and accurate. Context: I just finished a call with Priya Nair, office director, at Calloway Dental Group, a made-up clinic chain. We sell scheduling software for appointment-based businesses. These are my raw notes, which may be incomplete: three locations, front desk still uses paper + shared calendar; double-bookings weekly, hates it; wants SMS reminders, asked if we do those; worried about staff learning curve; their IT guy Marcus has to ok anything; i said i could probably get them a price this week; next: send pricing + demo invite, she said thursday works. Task: Turn the notes into a CRM update with these fields: summary (two sentences), what they need, objections or concerns, other people involved, next step, next step date, and open questions. Use only what the notes say. If a field has no information, write "not discussed". Do not set a deal stage, a close date or an amount. Flag anything in the notes that sounds like a promise we made. Format: Field name, a colon, then the entry, one field per line, ready to paste into the CRM. No bullets inside fields. After the fields, one line starting "Promises to check:".
What a good answer looks like
- Summary mentions three locations, paper plus a shared calendar and weekly double-bookings.
- "Other people involved" lists Marcus in IT as a required approver.
- Budget shows "not discussed" instead of an invented figure.
- "Promises to check" catches "get them a price this week".
How to check it before you use it
An AI draft can sound right and still say something you never told it. Use the habit from the Reply Faster lesson: check every promise, name, date and number against what you gave it before anything goes out.
- Compare "next step date" with your calendar. "Thursday" needs a real date, and the model may not know today's date.
- Make sure "not discussed" appears where you truly didn't talk about it, like budget or timeline.
- Set the deal stage yourself. The prompt deliberately leaves it blank.
- Fulfill or correct the flagged promise in your next message to the buyer.
Why this prompt works
Short, structured entries get used. Fields that read "not discussed" also tell the next person on the account exactly what to ask on the following call.
Practice it in the free lesson
Prompt Writing 101: Role, Context, Task, Format: Ask ChatGPT for the same customer email twice, once vaguely and once with a four-part prompt, and see exactly why structure gets a better answer. You do every step yourself in a practice copy of ChatGPT, and nothing touches your real accounts.
Also useful: Build an n8n AI Agent That Writes Deal Follow-Ups · Summarize a Meeting Transcript with ChatGPT
Terms used here
Related prompts
- Customer call notes summary prompt: Turn rough phone-call notes into a clean customer record: what they reported, what you promised and the next step, plus a short…
- Lead qualification prompt: Sort leads into Hot, Warm and Nurture with your own written rules. This ChatGPT prompt quotes the words behind each rating and…
- Sales pipeline review prompt: Paste your open deals and get a pipeline review for your sales meeting: stalled deals, missing next steps, past-due close dates…
More sales prompts
- Cold outreach email prompt: An honest cold email prompt for ChatGPT: verified facts about the prospect only, who you are and why you're writing, one easy…
- Competitor comparison one-pager prompt: Build a fair one-page comparison with a competitor using only facts you paste in, with every claim traceable and gaps marked…
- Demo request reply prompt: Reply to a demo request fast with a ChatGPT prompt that shows you read it, answers questions only from your product facts and…