In most scenarios, average is more than enough. To get better than average you need a process that works based on real expertise for ChatGPT to follow.

ChatGPT is a language model that predicts the most likely next word in a given sequence.

If you ask it to write an article on running shoes, it will offer up the most common phrases and sentences related to running shoes.

While it’s technically predicting tokens, which are more like syllables than words, and you can adjust something called temperature to make it less predictable, the point remains.

This is great for people who have less knowledge or experience than your average person on a topic. For instance, if you don’t know how to write a sales letter, ChatGPT can instantly write you an average one. It can write you an average email, tell you an average story, and brainstorm some average ideas. And in most cases, average is enough.

However, if you want to produce an excellent sales letter, relying solely on ChatGPT won’t be enough. You need a process for ChatGPT to follow—a process based on real expertise that works. You need to understand each step involved and what makes the output good or bad at each stage.

With a good process, ChatGPT can significantly speed up your workflow. It can also help you delegate complex tasks to other people. As a result, there’s now a huge incentive to translate tacit knowledge into explicit processes. If one or two people in your business know how to do something well, getting them to turn their process into a series of well-formed prompts means that everyone on the team can now do that one thing better than average.

Why would I ask ChatGPT to write me an average sales letter when I could use the sequence of prompts our sales team put together based on all their years of sales experience?