An AI Experiment to Help Buyers Ask Better Questions (and Feel More Certain)

Here’s one experiment I’m running to remove uncertainty for buyers.

The Observation

When I look at the deals that close fastest and cleanest, there's a common thread:
The buyer asked really good questions.

Not just “What does it cost?” or “How long does implementation take?”
But questions like:

• What problems do most of your customers come to you with?

• How were they solving them before?

• What metrics do you impact most across your customer base?

These are the kinds of questions that signal something deeper:
The buyer is trying to build a compelling internal story, one that survives scrutiny from legal, finance, security, and maybe even the CEO.

And I’ve noticed something else…

Different types of companies ask different types of questions.

• Enterprise buyers ask things like: “How do we ensure users only take actions we approve?”

• SMBs ask: “How much time will this save my reps?”

It makes sense, their risks, goals, and constraints aren’t the same.

So I started wondering:

What if I could give buyers a head start by surfacing the questions people like them asked — before they even think to ask them?

The Goal

Remove uncertainty.
Build trust.
Help buyers tell better stories internally.

How? By giving them the questions their peers have already asked, along with thoughtful answers at the moment they need them most.

The Experiment

I’m starting with the discovery phase (the first two calls).
Here’s what I do:

Step 1: 

Export a list of my customers using Scratchpad Views export feature.

Step 2:

Ask ChatGPT to enrich the list with data like industry, company size, revenue, and product offering.

Prompt:

I’m going to give you a list of company names. Please enrich the list by adding the following details for each company:

1. Estimated employee size

2. Estimated annual revenue

3. Industry

4. Sub-industry (if applicable)

5. Brief description of the product offering

Format your response in a table with columns:
Company Name | Employee Size | Estimated Revenue | Industry | Sub-industry | Product Offering

Here’s the list of companies:
[Insert company names here]

Step 3: 

Give ChatGPT the domain of a new prospect and ask it to find the similar customers from the enriched list.

Prompt:

I’m going to give you the domain of a company. Please find 2-3 similar companies from the enriched list above based on the following attributes:

• Employee size

• Estimated revenue

• Industry and sub-industry

• Product or service offering

Here's the domain: [insert domain here]
Please return a list of similar companies with a brief explanation of why each one is similar.

Step 4:

Pull the transcripts from the first two calls using Scratchpad’s export transcript feature with the list of matched customers.

Step 5: 

Ask ChatGPT to extract and categorize all the questions the customer asked and the answers given.

Prompt:

I’m going to provide you with several call transcripts from first and second sales calls with customers who are similar to a new prospect.

Please do the following:

1. Extract all the questions asked by the customers.

2. Categorize the questions by topic (e.g., pricing, implementation, integrations, ROI, etc.).

3. Include the corresponding answers given by the sales rep for each question.

4. Remove redundant or duplicate questions, combining similar ones into a single clear version with a comprehensive answer.

5. Format the final output as a guide titled:
   "Questions Customers Like You Asked — and How We Answered Them"

Each section should include:

• Category

• Customer Question

• Answer

I’ll now upload the transcripts.

Step 6: 

Review list and make sure everything accurate

Step 7: 

Package the questions and answers into a short PDF and proactively share it with the prospect.