The Holy Grail of GPT for Sales
As a Sales Executive getting to know your client is everything. This is why the best Sales Executives often will spend a lot of time investigating their prospects, to ensure they can talk about their particular needs instead of throwing around the common generic rubbish.
If I have a restaurant, I don't care about how a ChatGPT-based chatbot can increase customer engagement for instance. I want my website visitors to come to my restaurant, so I want to be explained how a chatbot can make more customers book tables, and not generic stuff explaining "how to increase sales with ChatGPT".
This implies that as we approach our prospects we need to tailor our message to our prospects according to their particular needs, and not throw around random marketing rubbish intended to catch "everybody". Hence, if we can create a GPT that somehow knows the most important parts of our prospects, and combine that with our products and services, we can create a tailored sales message, explaining the benefits for each individual prospect.
Well, we just did exactly that 😁
If you're in B2B, your primary tool of choice is probably LinkedIn InMail messages and direct emails. Your problem is that you don't have the time to spend 2 hours investigating each of the prospect you're sending emails to. The solution is to connect your own GPT with the ability to scrape your lead's website, and combine that with your own product and service offerings, to create a personalised and targeted sales message, speaking directly to the client's needs.
The best sales executives don't have answers, they have questions. When they've got answers to their questions, they can respond to needs with their own products and services. Selling is about asking questions, not giving answers
First watch the above video, then create two API endpoints. One for retrieving context data and another for scraping your prospect's website. Below is the code for scraping your prospect's website. Create two files in for instance a "sales-module" folder in your cloudlet.
.description:Scrapes the specified URL and returns the result to caller as structured result, with one segment for each section found.
Then we'll need an endpoint for retrieving context data.
.description:Returns context from your training snippets as a string, using VSS search through your training data.
// TODO: Change the [type] below to your machine learning type in your cloudlet!
The first file above will scrape the client's website, and the second file will retrieve context data from your machine learning model. Make sure you change the [type] to your model's actual name.
Creating your GPT
Now the only remaining thing is to create a GPT and add an instruction to it as follows;
You are a marketing expert working for YOUR-COMPANY-NAME-HERE. Your task is to generate highly personalised content for individual messages, such as emails and LinkedIn InMail messages. You will do this by scraping my lead's website, for then to retrieve context, and create a highly personalised message using keywords and constructs from the lead's website, his or her name, his or her title, in addition to relevant information about products and services found in the context. If I ask you for incomplete data, then ask me questions until you have everything you need to perform your task. You will need the following information to perform your task:
- The lead's name
- The lead's title
- The lead's website. If I supply you only a URL without a scheme, assume the website is served over HTTPS
Once you have all of the above, scrape the lead's website using the scrape-website action, retrieve relevant context using the get-context action, and return a message according to my instructions. Following these rules as you construct the message:
- If you need to sign the message, sign it with YOUR-NAME-HERE, YOUR-TITLE-HERE at YOUR-COMPANY-NAME-HERE
- Use keywords, constructs, and ideas for the content of the message found as you scrape the lead's website
- Use "PRODUCT-YOU-ARE-TRYING-TO-SELL" as your keyword and add 1 to 5 primary keywords found when scraping the client's website as you invoke the get-context action to make sure the action returns relevant and personalised context
- Show me what keywords you were using as you invoked the get-context action, and allow me to change them if I want to
- Make the message highly personalised to the lead's title, name, and the content found as you scrape his or her website
- Use all relevant keywords and constructs you can imagine as you construct your message found as you scrape the lead's website to reduce the risk of having the message being labelled as spam
Then all you need to do is to connect your GPT to your API endpoints, and your custom GPT will be able to create highly personalised sales emails and messages. Below you can see a screenshot of me interacting with our own Sales GPT. Notice, to understand the value you have to watch the video above.
By using this method you get the following advantages.
- Highly unique and personalised emails and messages
- Low spam score on your messages since your messages and emails will use the same keywords found in email conversations your prospect is sending and receiving every single day himself
- The ability to pitch your product and your services to specific needs your prospect have
Basically, your emails are no longer generic random marketing rubbish emails, but highly tailored and unique, directly targeting your prospect's needs. If you need help with creating your own Sales GPT, we would love to help you out. You can contact us below to get started.