AI for pricing analytics. I tried very hard to stay away from the topic because I didn’t want to add to the noise. But I read so much nonsense every day that I cannot stay silent anymore and I feel it’s my responsibility to bring some level-headedness to the discussion. And hopefully, I won’t be adding to the noise.
Over the next few weeks, I will be discussing what using AI for pricing analytics really means: where it’s useful, where it’s not, how to properly use it.
Before we start, let me make two important comments:
- I’m what you could call an AI-skeptic. I was a very late adopter and I still have some reservations about it. However, I’ve also slowly started to see its value and I am now using it more and more in my work with clients. All this to say, don’t expect me to talk about how it’s going to revolutionize your business and solve all the problems in your life. It’s just another tool at your disposal, and like every tool, it can help you a great deal if you use it properly…but it can also be very dangerous if you don’t.
- AI for pricing analytics, or for data analytics in general, is not just about crunching the numbers. If you think you have become a great pricing analyst by simply loading your data into ChatGPT and asking it a bunch of questions, think again. There is a lot more to having a best-in-class pricing analytics function than just looking at the data. I know that from spending the last 15 years building world-class pricing analytics capabilities for my clients…and for myself.
That takes me to today’s post: if your goal is to build world-class pricing analytics for your business, what do you need?
As the diagram below shows, there are five pillars to pricing analytics (and, really, to any data analytics initiative):
- Culture: building habits of making pricing decisions based on data, not feelings, is key to driving profitability.
- Processes: going from a business question to number-crunching and back to a business decision is not something you can improvise, it’s a process.
- Data: having quality data is obviously a requirement for pricing analytics but you don’t need much to get started.
- Skills: knowing what questions to ask the data and how but also knowing how to respond to practical business questions, is what makes a great analyst.
- Technology: having the proper tools makes pricing analytics much more efficient and effective (so please, forget Excel).
Yes, AI can help you with all five, but that’s not really the point. All five pillars require a commitment from the business’ leadership, just like it does for any other function within the organization.
This framework also still requires “humans in the loop”. Sorry but you won’t be able to entirely outsource your pricing analytics to an AI agent, at least not yet. There are many reasons for it but to give only one: control and transparency. Unless you are perfectly comfortable with your pricing analytics being turned into a complete black box (and you shouldn’t!), you still need a (human) team to look into what the AI did and verify its output before pushing prices on to customers.
Stay tuned over the next few weeks (probably even months) as I unpack each pillar in detail and show you how AI can help you without jeopardizing your pricing.
