Jimmy Kim argues that having data and tools isn't enough for personalization. Without proper structure, bandwidth, and testing, teams drown in thousands of branching rules they can't maintain. AI changes this, but only if your data is clean and your testing layer is solid.
segmentation, we absolutely have the data, we absolutely have the tooling, but we've seen through audits and through working with a lot of brands hundreds of these accounts, and it's like even if you have the data, it doesn't necessarily mean you have the structure or the bandwidth necessarily to maintain it. And that goes for both agency and in-house teams because there's so much logic that you have to maintain and so much static logic that if you want to be truly personalized, you're either doing a lot of testing, very niche kind of branching logic, or to your point, very niche segmentation. But then even at like a massive scale, you're still not able to do this any single person or any single team tens of thousands of times per month. So that's also where the unlock from AI comes, where it's like there are a lot of capabilities that if the data is structured properly and if the testing layer is there, you can really supercharge these learnings with the tooling that's now available that just 5 years ago wasn't. So there's also like a huge profound shift happening in how personalized you can get. But you do need to do that through some sort of like rigid framework or like some robust testing framework. Otherwise you sort of run into like, I have so much data, but I don't actually know what to do with that data.
