Part 1: AI Prompting Basics for Sellers

Prompt engineering will be one of the must learn skills for modern sellers in 2025. Unfortunately, many don’t even know the basics.

‘Prompt engineering’ sounds technical and unapproachable for non-technical folks, but it’s actually easier than you’d think. After all, it’s just tweaking how you write in natural language to make it easier for ChatGPT to deliver on what you want. Here’s an example of going from bad to good:

Sellers who learn these techniques will be far more successful leveraging AI in their day-to-day.

These are some of the basics that I’ve learned while writing prompts for customers at Clay.

WTF is Prompt Engineering?

Prompt engineering is the process of creating and refining the “prompts” (i.e., the text or instructions) used to interact with large language models (LLMs) so that you get the most useful and accurate responses.

By carefully crafting prompts—stating the goal, providing necessary context, specifying the desired format, or even giving examples—you guide the model’s output. Effective prompt engineering leads to higher-quality, more targeted answers, and reduces the need for extensive post-processing or iterative follow-up queries.

Input Options

A prompt can be accompanied by additional inputs. For example, you could upload a call transcript or CSV of data that you want to analyze.

These are some basic ones that we can use as sellers:

1. Call transcripts

2. CRM data exports

   • Activities

   • Pipeline

   • Closed Lost deals

   • Closed Won deals

3. Your case studies or marketing material

4. Website links of your customers

5. 10-K reports

6. Quarterly earnings reports

7. RFPs

8. Proposals

9. Call notes

10. Emails

Data Analysis Example with a CSV

This is an example of how I analyzed a CSV data set of subscribers for my other newsletter claymation (GTM Engineering with Clay and AI).

The data upload: CSV file of subscriber emails + job titles.

The prompt:

The output: