Retail personalization, in plain English.

Ideas from the merchandising floor and the data pipeline.

Purchase Signals vs Click Signals

Purchase Signals vs Click Signals: What Your Shopper Data Is Actually Telling You

Clicks tell you what caught someone's eye. Purchases tell you what they actually needed.

8 min
How Mid-Size Retailers Can Compete on Personalization

How Mid-Size Retailers Can Compete on Personalization Without Amazon's Data

Amazon personalizes because it has billions of purchases. You have thousands. That is actually enough, if you use them right.

6 min
The Merchandiser's Algorithm

The Merchandiser's Algorithm: What AI Should Actually Learn from Your Buying Team

Your best merchandisers make decisions based on what customers have bought before, what they returned, and what they keep coming back for. That is exactly what a good personalization model should learn too.

9 min
Cart Abandonment Is a Discovery Problem

Cart Abandonment Is Not a Checkout Problem, It's a Discovery Problem

Most abandoned carts happen before the checkout page. The shopper left because the products on the category page were not relevant to them. Fixing your checkout flow will not solve that.

6 min
Building Storefronts for Seven Retailers

What We Learned Building Personalized Storefronts for Seven Retailers

Our first cohort of early-access merchants came from four different retail verticals. Here is what the data patterns had in common and where they diverged.

10 min
The Hidden Revenue in Product Sorting

The Hidden Revenue in How You Sort Product Category Pages

Default alphabetical sort is a placeholder, not a strategy. Sorting by purchase likelihood per shopper segment can shift add-to-cart rates.

7 min
One Store Per Shopper Framework

One Store Per Shopper: A Framework for Thinking About Retail Personalization

The end state for personalization is not a slightly better homepage. It is a different storefront for every visitor.

8 min
Why Purchase History Outperforms Browse Data

Why Purchase History Outperforms Browse Data for Product Recommendations

Browse data is noisy. A shopper might click on ten items out of curiosity and buy one. Training a model on clicks trains it on distraction. Purchase data trains it on intent.

8 min
The Merchandiser's Toolkit in the Age of AI

The Merchandiser's Toolkit in the Age of AI: What Changes and What Stays the Same

AI does not replace merchandising judgment. It amplifies it, if the model is designed to encode the right inputs from the start. Here is what that actually looks like in practice.

6 min
Session Depth and Conversion Depth

Session Depth and Conversion Depth: Why They Move Together

Shoppers who browse deeper buy more. But depth only happens when the products on each page feel worth exploring. Relevance drives depth, depth drives conversion.

7 min
Running Our First Retailer Pilot

Running Our First Retailer Pilot: What We Got Right and What We Had to Rebuild

Our first pilot was a home goods retailer with a strong repeat-buyer base and a storefront that showed the same top-20 products to everyone. What we found in the data surprised us.

9 min