This content is locked. Please login or become a member.
Focus on business impact
In a nutshell, here’s how to sell AI. If you’re talking about generative AI, something like chatGPT, a large language model, it’s really, really easy to sell. It seems so amazing. It’s potentially a panacea. According to all the headlines, it’s on the brink of solving all business problems automatically with the slight side effect of displacing huge amounts of workforce. I believe that that’s hype, and I think the important thing for selling generative AI is to not buy into the hype, but instead be realistic. Focus on a credible value proposition. We’re going to help customer service agents by drafting their response to customers. Make it credible, concrete. Use existing technology. Don’t expect the thing to suddenly make a breakthrough as if human-level general intelligence is knocking on the door.
Whereas predictive AI, which is older but not old-school, very much still has great amounts of untapped value. You have to push a little bit harder. It doesn’t sell itself so readily, but it is oftentimes more valuable because it has the potential to improve your largest-scale existing operations. But the key that’s often missed is don’t lead with the technology. We love the technology, but your stakeholder, your client, the executive, the people that need to be convinced, the people who run the large-scale operations that are potentially going to be improved with the model, they care about very practical business situations and value propositions. Exactly what is this going to predict, and then how is it going to be used. So start with the deployment goal and why it may be valuable. Create a back of the napkin scratch arithmetic that shows just how much it’s going to improve key performance indicators, bottom line metrics, and also describe how you’re going to mitigate risks because it does involve change. It’s a big change. This is the largest-scale operation, the most to potential gain by improving it, but also the highest risk by changing it. This is how the company is making money, for example. This is how a hospital is saving lives. It can be very tricky to change it. You can talk about risk mitigation in how that deployment’s going to take place, incrementally, for example.
If you can keep the value proposition as short as possible, you stand to actually get through the pitch before losing the focus and concentration of your audience. And then open it to questions. Because you have no idea which aspect of the project is going to be potentially questionable or that they’re going to be skeptical about or have questions about. Just keep on point the value proposition, and then let them lead the degree to which you get into the details of the core technology.
Deliver a visceral experience
UPS dramatically improved their delivery efficiencies of millions of packages by predicting tomorrow’s deliveries. But this system was very new and innovative at the time. It meant big change. And the leader of the project, Jack Levis, had to sell it to one of his executives that sat above him at the organization. And that executive’s name was Chuck. So he presented to Chuck. He did the slides. He stuck the landing, and Chuck looked at him like, okay. Not interested. And it turned out what he ended up doing to help sell it to the executive is literally taking the executive for a ride. So he used this delivery optimization system, and they got into a car, pretended it was a delivery truck, and followed the instructions for delivery and driving route from the system. And then this light bulb goes on over Chuck’s head, and he’s like, this kind of thing could really stand to deliver an amazing bottom-line impact. And that’s what made the difference. By actually delivering a visceral experience of the value rather than just sort of some numbers on a screen, That’s how Jack sold the project initially to executives.