How big is your baseline (control group) when you do dynamic pricing?
We keep 10% to 20% of traffic set aside - no artificial intelligence (AI) - so that we know how well our AI is doing. Say, for example, your price is $30. Well, 10% to 20% of your shoppers will get that price. We compare the baseline’s performance with the performance of the shoppers that got the AI. Which one made more money? How much more did it make? That tells us the value the AI is creating.