TDLR: I used ChatGPT to build a use case for my prospect. The tool made things up and my prospect found out. 4 month deal down the drain.
If you haven’t used ChatGPT yet as a sales person, it can do wonders. Welcome to 2023. You can use this free tool to learn about competitors, analyze 10ks, and write many of your emails. But what if you took it to the next level and had it help build out use cases? Don’t, it killed my deal.
A Pipeline Press reader sent us his story about how 4 months were wasted because ChatGPT had a hallucination. An AI Hallucination is when the AI tool like ChatGPT, generally known as a Large Language Model (LLM), essentially makes things up. News this past summer of multiple attorneys using ChatGPT to write case arguments and the tool cited non-existent court cases, the judge found out and the attorneys were fined $5,000.00.
Our Pipeline Press reader, we’ll call him Steve, works for a small startup with not much marketing but a great product. The sales team is small and scrappy that needs to think creatively to get deals across the finish line. In comes ChatGPT, where Steve needed help building a specific use case for a prospect that was in the final stages and in commit for Q4.
Unfortunately for Steve a senior member of the procurement team questioned the validity of the project and if it could be technically done. This new blocker requested use case documentation from other customers of Steve and his company, however with no marketing team and no use case evidence Steve decided to think on his feet and use ChatGPT. Although the companies senior leadership spoke with Steve’s champion and the senior procurement officer, a use case and documentation was still requested.
Using ChatGPT, Steve put together a technical Use Case that would surely pass the sniff test. After careful review and approval from his director the Use Case was put in a pretty slide deck and sent off. Unfortunately for Steve, ChatGPT had a hallucination and created a technical impossibility in the documentation that the prospects IT team identified and called out.
The prospect then requested to speak with the “ChatGPT customer” as a reference. Steve was now in a world of confusion on what to do. Should he go to his management and share that his prospect doesn’t believe the documentation and that it was made up by ChatGPT or should he say the customer doesn’t feel comfortable speaking about their technical roll out? Steve chose the latter.
The prospect with no where to go, and the feeling of trust being broken, decided to pass on things and remain in status quo.
