Chat-GPT Has An Answer for Everything. But Are Your Buyers Asking the Right Questions?

Today, almost everyone from college students to professionals turn to AI chat bots as a first line of investigation when they have questions. From describing the plot of an assigned reading, to surfacing insights about the market, AI has an answer for, quite literally, anything.

So it makes sense that buyers turn to it when they have a problem or need a solution, too. Answers from an AI chat bot are fast, coherent, and easy to understand. Unfortunately, just because the answers are quick doesn’t mean they’re right. Worse, they can be downright dangerous. And, worst of all, buyers may be asking the wrong questions to begin with.

The Danger Of Bad Questions

When my daughter adopted a pair of guinea pigs recently, she spent a great deal of time researching to make sure she was ready for them. When they arrived, she spent even more time making sure they were okay and tending to their needs. Soon after, she noticed them behaving strangely and turned to an AI chat bot to find out if it could mean something dangerous.

The chat bot said yes. What she was describing meant they could be very sick, and she should take them to a small animal vet immediately. I suggested she call the breeder first. The breeder told her that the behavior could indeed be mistaken for signs of illness in a different context, but was actually very typical for guinea pigs when they are first settling in and feeling nervous about their new home. The advise from an experienced breeder: “just relax.”

When my daughter turned to the chat bot, she didn’t realize she was priming its answer by asking “could it be dangerous.” In turn, the chat bot gave her the answer it thought she expected, not the grounded answer that would help her make good decisions.

Buyers, especially in complex sales, often do the same thing, and it might not send them to the animal hospital. But it will lead them to bad decisions.

The Problem Is More Widespread Than We Realized

Sachin O. wrote about this problem on LinkedIn recently, citing research from MIT, the University of California, and others, on an AI behavior called “sycophancy,” and its impact on users. Rather than seeking a “true” answer, AI chat by its very nature seeks an answer that will continue to elicit engagement. In most cases, that means giving the reader what they expect, not what is best for them.

Buyers need your salespeople to stop trying to sell, and become professional decision-making catalysts.

Engaging with AI chat bots without critical awareness reinforces the user’s existing beliefs, leading to over-confidence even when they are wrong. At its worst, this can lead to psychosis, a condition in which the human becomes so convinced of untrue “information” that they stop believing the world around them and live instead inside the psychosis.

In California, one psychiatrist reports having hospitalized 12 patients in one year for this phenomenon. Multiple lawsuits have been filed over this issue, indicating that it is widespread and growing.

Why Salespeople Have To Step Up

When a buyer in a complex B2B environment sets out to solve a problem, asking the wrong questions, or asking them in the wrong way, can lead to “sycophantic” answers from AI chat bots. While this isn’t likely to send them to the hospital (or even the vet), it very well can cause them to make bad decisions.

The trouble is, buyers responsible for executing a complex purchase in a B2B context rarely make the same purchase more than a few times in their career. How often does a company purchase a new CRM, after all? And because they don’t make the decision very often, they don’t know what they don’t know.

They might ask, “What is the best CRM?” and a chat bot will list some of the most popular platforms, or the ones that have paid for the privilege of being noticed by LLMs, but that doesn’t mean they’ll be the best solution for that particular buyer’s needs. And the buyer won’t know.

They only know what their friends have told them, what they’ve read about on the internet, and what they assume to be true. If they take these assumptions to an AI chat bot, it may offer some valid suggestions based on those criteria… but if the criteria are based in bad assumptions and unclear distinctions, the outcome is going to be bad, too.

Professional salespeople, however, solve the same problem repeatedly. They are experts in that problem. Our sales teams know what questions to ask to help buyers understand what they need from their CRM solution. Your sales teams know what questions to ask to help buyers understand what they need from your solutions, too. Or, they should.

Because in the age of AI, buyers don’t need old-fashioned salespeople. They need guidance. They need helpers who understand the terrain, can push back on their bad assumptions, help them uncover what they don’t know that they don’t know, and put the right questions in their hands so they get answers that actually help them.

In the age of AI, buyers need your salespeople to stop trying to sell, and become professional decision-making catalysts. That is what will set your team apart from the rest.