Axes

In my newsletter the other week, I spent some time talking about creative volume and how it fits with customer development.

One of the things that’s happened since then is ChatGPT’s 4o model update, which (as the DTC chatter goes) effectively brings the cost of ad creation for brands down to zero—and, potentially, scales the amount of ad volume to unthinkable levels.

That last newsletter feels maybe more relevant in the wake of that change, but, as Adam wrote last week, we spent a lot of time at Shoptalk and DTSki talking to brands about the idea of personalization. And I think this change is actually very relevant to those conversations. 

From Adam last week: 

It’s not that people are opposed to it or even scared of it; they’re just not thinking about it too much.

This has been on my mind. 

When Adam and I left DTSki last weekend, we turned around and headed to a company offsite, where we’ve spent even more time talking about this. (So, forgive me if this week’s newsletter feels a little disjointed; I’ve been in four different timezones in the last seven days.)

I think it helps to visualize personalization as living on two axes: product and content.

Your product is either relevant for a consumer or it’s not. It’s not quite that binary, but the product(s) you choose to sell effectively segment the market for you. Layer on quality and price, and you’ve segmented to a pretty narrow slice of the consumer market (even if it’s still a pretty massive number of people). With a store with a large catalog, determining which products are going to be relevant for a given customer is one of the largest (and most commonly implemented) personalization types. 

Content, mostly, is a way to sell that product. It also can also pull people into the product category if you can tell a novel story about how they can use it / why they need it. 

Personalization here is basically being better at sales, and finding ways to tell the right story to a given customer that convinces them to give the product a try. Traditional brand building has often taken a quasi-consecutive approach to these axes: Once you have a hero product, you first build your brand for one consumer type via content. Then, as you saturate that consumer type, you choose whether to iterate on product (one axis) to sell more to your existing customer or on content (another axis) to find a new angle so that you can sell the same product to new customers. 

Often a brand will do a bit of both types of “expansion”, and, over time, that brand becomes known for a market position, because it personalized the brand to a specific consumer.

This approach has naturally flowed to ecommerce, but now there’s the opportunity to have the site reflect the depth/breadth of those axes. So, you end up with product-related personalizations (product recommendations) and content-related personalizations (perhaps showing a specific model type to returning customers). And, because of technology, you can also move past the consumer (a broad term for a group of similar people) to the customer.

This is one of the main advantages of ecommerce. The fact that you are digital lets you rebuild the store for everyone who comes in, and lets you know things about those customers!

That is such an advantage over brick and mortar, where doing personalization is like driving a cruise ship. REI can have a totally different assortment at their store in Denver vs SoHo, but they can’t change that on demand for every customer who comes in. 

I think we’re going to enter an age—as this tech all evolves—where this rapid store reconfiguration is going to be the norm for ecommerce. It’s the natural advantage of selling online, and the best brands are going to use it. Brands will be more quickly expanding along both axes—product personalization and (especially) content expansion—with the AI tools available to them. 

The logical conclusion is a completely generative experience for each customer who comes to the site. I don’t know quite how far away that is, but I expect we’ll see a lot of progress here in the coming years. 

But then I go back to the conversations with brands from the last few weeks where they’re already struggling to implement customer-level personalization in a scalable way. Will they keep up? Maybe the current blocker is effort and labor, which AI could solve. We will see.  

As you imagine more and more generative personalization, it also begs the question of what a “brand” actually becomes. If a brand used to mean the same thing to diverse sets of people, but now it means different things to even more diverse sets of people, then how do you build an experience that keeps a through-line to the purchase? 

I’m starting to turn this over in my head, and I’m wondering if brands are, too.