A guide to your first biz dev custom GPT
This resource is based on The AI outreach stack that handles prospecting, follow-up, and scheduling, featuring Amanda Pressner Kreuser of the Masthead, published on the AI Lab by ActiveCampaign.

Get the playbook
Amanda Pressner Kreuser built five custom GPTs for her entire biz dev workflow, but you don’t need all five to start seeing results. This quick-start guide walks you through building one—the follow-up email GPT—which delivers the highest time savings for the least setup. The full post covers the complete five-agent system when you’re ready.
What you’ll build
A custom GPT that drafts personalized follow-up emails after sales calls. You feed it a call recording or notes, and it generates a ready-to-edit email in your voice. Amanda says this cut her follow-up time from 15–20 minutes per email to about two minutes.
Time commitment
- Quick version (this guide): 30–45 minutes
- Full version (see original post): 4–6 hours for all five agents
What you need
- ChatGPT Plus account (custom GPTs require a paid plan)
- 3–5 of your best follow-up email examples (past emails you’ve sent that landed well)
- One recent call recording or set of call notes to test with
The build (minimum viable version)
Step 1: Gather your email templates—10 minutes
Pull 3–5 follow-up emails you’ve already sent that represent how you want to sound. Save them in a single document with labels like “post-discovery call,” “post-intro call,” or “re-engagement.” These become the voice reference for your GPT.
Quick version: If you don’t have 3–5 saved, grab your 2 most recent follow-up emails from your sent folder. Two examples are enough to start.
You’re doing it right if: You have a document with at least 2 labeled email examples that sound like you at your best.
Step 2: Create the custom GPT—15 minutes
Open ChatGPT and ask it to help you write the system prompt:
Custom GPT
I want to set up a custom GPT to draft follow-up emails after sales calls, and I want you to help me create a prompt for this custom GPT so I never have to write these same instructions again.
My tone is [YOUR_TONE]. I work at [YOUR_COMPANY], which does [WHAT_YOU_DO]. I’ll upload my email templates as reference.
Answer the clarifying questions ChatGPT asks. Then copy the generated prompt, go to “Create a GPT” in the ChatGPT sidebar, paste the prompt into the Instructions field, and upload your email template document as a knowledge file.
Quick version: Don’t overthink the prompt. Amanda’s approach is to get it roughly right, then refine after testing. If the first attempt produces bad output, delete the GPT and start again with clearer instructions.
You’re doing it right if: You can see the new custom GPT in your ChatGPT sidebar and it responds when you type “I just had a call with a marketing director at a mid-size company. Help me draft a follow-up.”
Step 3: Test with a real call—10 minutes
Take a recent call recording or your notes from a real meeting. Paste them into the custom GPT along with the person’s name, company, and what stage of your sales process they’re in. Review what comes back.
Quick version: If you don’t have a recording, type up 3–4 bullet points from your last sales call—who you met, what they need, and what you discussed. That’s enough for a useful test.
You’re doing it right if: The GPT produces a follow-up email that sounds roughly like you, references specific details from the call, and needs only light editing before sending.
You’re done when…
You have a custom GPT that takes call notes as input and produces a follow-up email draft that sounds like your writing. You should be able to edit the output in under five minutes instead of writing from scratch.
What to build next
- Build a Referral Scout GPT to filter LinkedIn contacts against your ICP
- Create a Scheduling Assistant GPT that suggests meeting windows from your calendar
- Compile a full library of 10–15 email templates and upload them to an expanded Email Library GPT
- Set up a Google Apps Script to automatically log outreach activities to a tracking spreadsheet
Ready for the full story?
Read The AI outreach stack that handles prospecting, follow-up, and scheduling, featuring Amanda Pressner Kreuser of the Masthead, published on the AI Lab by ActiveCampaign.
Or, if you’re ready to build, here’s:
- A step-by-step checklist to build your own biz dev agent workflow and
- The actual prompts Amanda used to create her 5 agents that run her prospecting and business development
Related
More data from the AI Lab.