Most coaches who struggle on discovery calls have the same problem: they walk in underprepared and walk out with a vague promise from the prospect to think about it. Then they send a generic follow-up email two or three days later and the conversation dies.
The fix is not more charisma or a different pricing structure. It is preparation — and specifically, a preparation system you can run in under thirty minutes before every call using AI tools you already have access to.
This article gives you that system. Four steps, four AI prompts, one complete pre-call and post-call workflow. Every prompt in this piece is copy-paste ready. The only inputs you need to add are your own details and the prospect’s information.
1. Why Most Discovery Calls Fall Apart Before They Start
A discovery call is a sales conversation. The goal is to diagnose a problem accurately, present an offer that addresses that problem specifically, and close toward a decision. Most coaches treat it as an informal conversation and then wonder why the conversion rate is low.
The three most common preparation failures:
- No prospect research. Walking into a call without knowing anything about the person beyond their booking form response. The questions feel generic because they are.
- No tailored script. Using the same opening questions regardless of who the prospect is or what they said they needed. The conversation feels scripted in the bad way — rigid, not responsive.
- No follow-up system. Improvising the follow-up email from memory hours later, resulting in a message that misses the specific thing the prospect said they needed and reads like a template they have seen before.
AI does not fix a bad coach. But it eliminates all three of these preparation failures in under thirty minutes. Here is the system.
2. The Four-Step AI Preparation System
Step 1: Research the Prospect
Run this before every call. Takes five minutes. Requires whatever information you have about the prospect — their booking form response, a LinkedIn profile, a website, or even just their job title and the problem they described.
Paste everything you know into Claude or ChatGPT and run the prompt below. The output gives you a working hypothesis about what this person actually needs, what they are afraid of, and what is most likely to stop them from saying yes.
I am a [your niche] coach preparing for a discovery call with a prospect. Here is what I know about them:
Name: [name]
Background: [paste LinkedIn bio, website about page, or any relevant detail]
How they found me: [referral / social media / search]
What they said in their booking form: [paste their intake form response]
Based on this, give me:
1. The 3 most likely problems they are trying to solve
2. The 2 most likely objections they will raise on this call
3. The single outcome they most want from working with a coach
Do not skip this step even when you think you know the prospect well. The AI synthesis will surface something you missed or frame the situation in a way that sharpens your questions. The five minutes it takes to run this prompt will change how you open the call.
Step 2: Build a Tailored Script
Generic discovery call scripts fail because the prospect feels the template. A tailored script — built around what you know about this specific person — allows you to ask questions that make the prospect feel genuinely understood rather than processed.
Take the research summary from Step 1 and feed it into the script prompt below. The output is a complete 30-minute script with sections you can adapt on the fly. Use it as a guide, not a teleprompter — the structure keeps you from missing the critical elements, but the conversation should still feel natural.
Write a 30-minute discovery call script for me. I am a [your niche] coach. My offer is [brief description]. My price point is [price].
The prospect I am speaking with is: [paste the research summary from Step 1]
Structure the script in 5 sections:
1. Opening (2 min) — build rapport, set the agenda
2. Discovery questions (10 min) — 4 open questions tailored to this prospect
3. Diagnosis (5 min) — reflect back the problem clearly
4. Offer presentation (8 min) — connect my offer to their specific problem
5. Close (5 min) — direct ask, handle the two likely objections identified above
This prompt is best run in Claude. If you want faster variation volume — multiple script versions to compare — run it in ChatGPT instead. For the definitive breakdown of which tool performs better on which coaching tasks, see ChatGPT vs Claude for Coaches.
Step 3: Prepare for Objections
Objections are not surprises. Every prospect at a given price point and niche has a predictable set of concerns. The difference between a coach who handles objections well and one who stumbles is not natural talent — it is preparation.
Run the objection prompt below before every call. Read through the responses once. You will not memorize them and you are not supposed to. The goal is to have thought through each scenario so that when the objection arrives mid-call, you have a direction rather than a blank. The language comes out more naturally because you have already rehearsed the logic.
I am a [your niche] coach. My offer costs [price]. I have a discovery call today with a prospect who [brief description from Step 1 research].
Generate 5 realistic objections this prospect might raise and write a calm, non-defensive response to each one.
Objections should be specific to this type of prospect, not generic sales objections.
Each response should: acknowledge the concern, reframe without dismissing it, and redirect toward the outcome they said they want.
ChatGPT is the better tool for this step because it generates higher variation volume faster. Run the prompt and ask for five objections. If any of them feel off for this specific prospect, replace them with what you expect to hear and run again. The process takes less than ten minutes.
Step 4: Write the Follow-Up Sequence Before the Call
This is the step most coaches skip, and it is the one that costs them the most closed clients. The follow-up email written two days after a call from memory is never as good as one written immediately after the call from actual notes.
The system: use Otter.ai to transcribe the call in real time. Immediately after hanging up, paste the AI-generated summary into Claude and run the follow-up prompt. Send Email 1 within two hours. Set reminders for Emails 2 and 3 before you close your laptop.
I just finished a discovery call with a prospect. Here are my notes from the call:
[Paste Otter.ai transcript summary or your own bullet-point call notes]
Write a 3-email follow-up sequence:
Email 1 (same day, within 2 hours): Thank them for the call. Restate the core problem they described. Restate the outcome they said they wanted. Include a clear next step.
Email 2 (Day 3 if no response): Brief check-in. Add one piece of value relevant to what they said — a resource, a short insight, or a relevant question. No pressure language.
Email 3 (Day 7 if still no response): Close the loop. Make it easy for them to say no so they do not feel chased. Leave the door open professionally.
If you are not using Otter.ai to transcribe your discovery calls yet, that is the highest-leverage change you can make to your conversion process. The full breakdown of why is in The 5 AI Tools Every New Coach Needs.
The three-email structure matters. A single follow-up email puts all the pressure on one message. The sequence gives you three distinct opportunities to stay in front of the prospect with escalating specificity — first as a summary, then as a value add, then as a professional close. Coaches who use this sequence consistently close a higher percentage of calls that did not produce an immediate yes.
3. Putting It Together — The Full Pre-Call Timeline
Here is the complete workflow timed out so you know exactly when to run each step.
- 48 hours before the call: Run Step 1 (Prospect Research). Review the output and add any additional context you have gathered. File the research summary where you can access it quickly before the call.
- 24 hours before the call: Run Step 2 (Script Generation) using the research summary. Read through the script once. Mark the two or three questions you want to make sure you hit regardless of how the conversation flows.
- Day of the call, 30 minutes before: Run Step 3 (Objection Prep). Read through the five objection responses once. You are not memorizing them — you are loading the logic.
- Immediately after the call: Pull the Otter.ai summary. Run Step 4 (Follow-Up Sequence). Send Email 1 within two hours of hanging up.
- Day 3 and Day 7: Send Emails 2 and 3 per the sequence. Set the reminders before you close your laptop after the call.
Total active time for the full workflow: under thirty minutes across all four steps. The return on that thirty minutes — more confident calls, more personalized conversations, and a follow-up process that runs automatically — compounds with every call you do.
4. The Reason This Works When Improvising Does Not
Improvising discovery calls is a skill that develops slowly and expensively — through failed calls, lost clients, and months of inconsistent conversion rates. The system above compresses that learning curve by giving you structure that you can iterate on rather than starting from scratch before every call.
After ten calls using this system, you will have a library of research outputs, script variations, objection responses, and follow-up templates that you can refine. The AI is not replacing your judgment on the call — it is handling the preparation work so your judgment can operate at full capacity during the conversation.