The Conversion Intelligence Blog

Working notes on DTC collection page strategy, conversion rate versus revenue per session debates, product ranking models, and what the clickstream actually tells you. Written by the Cartlyzer team.

Latest articles

A/B test comparison of DTC collection page hero banners
Experimentation

What We Learned Running 40 Hero Banner Tests on DTC Collection Pages

Hero banner CTR is the wrong metric. After 40 A/B tests, the single most predictive metric was time-to-first-product-tap.

Dara Mensah
Bar chart showing top sellers vs top converters discrepancy
Merchandising

Why Your Top Sellers Are Not Always Your Top Converters

Sorting by units sold sounds logical. It's not. Static bestseller lists cost you when visitor intent doesn't match catalog rank.

Jennifer Nguyen
Hierarchy diagram of real-time behavioral signals for merchandising
Product Intelligence

Real-Time Signals That Actually Predict Purchase: A Practical Ranking

Hover time beats scroll depth. Add-to-cart beats wishlist. Here is our signal hierarchy based on eight months of live inference data.

Dara Mensah
Seasonal sales wave pattern overlaid on DTC product page strategy
Strategy

How DTC Brands Should Rethink PDP Strategy for Seasonal Peaks

Holiday peaks create very different visitor intent profiles. Here is how to adapt your PDP hierarchy without a full site redesign.

Luca Ferretti
Shopify Plus storefront optimization levers
Platform

Shopify Plus Storefront Optimization: The Levers Most Brands Have Not Pulled

Most brands use less than 30% of Shopify Plus levers for merchandising. Here is the priority order for brands scaling from $5M to $20M GMV.

Luca Ferretti
First-party session data signals replacing third-party cookies
Privacy and Tech

Personalization Without Third-Party Cookies: What DTC Brands Actually Have to Work With

The death of third-party cookies shifted what data is worth collecting. First-party session signals can do what cookies never could.

Jennifer Nguyen
Comparison of abandoned cart and browse abandonment patterns
Conversion

Abandoned Cart vs. Abandoned Browse: Why You Should Care More About the Second One

Abandoned cart recovery is solved. What most DTC brands don't recover from is browse abandonment: the visitor who spent four minutes on your collection page and left without clicking a product.

Jennifer Nguyen
Three ranking algorithm approaches shown as visual decision trees
Technology

Product Ranking Algorithms Explained: Rule-Based vs. ML-Based vs. Hybrid

Understanding three ranking approaches tells you exactly when to use rules, when to let a model decide, and when a hybrid gives you the best of both.

Dara Mensah
Two metric gauges comparing conversion rate and revenue per visitor
Analytics

Conversion Rate Is the Wrong Primary Metric. Use Revenue Per Visitor Instead.

A 4% conversion rate sounds good until your average order value drops 30% because the model optimized for easy conversions. Revenue per visitor surfaces this.

Jennifer Nguyen
Five common merchandising pitfalls for DTC brands
Strategy

Five Merchandising Mistakes DTC Brands Make When They Hit $2M in Annual Revenue

At $2M ARR, DTC brands typically hire their first dedicated merchandising person. That hire walks into a system that was never built for them.

Luca Ferretti
Behavioral prediction model activating before a shopper scrolls
Product Intelligence

Can You Predict an Add-to-Cart Before the Shopper Scrolls? Yes. Here Is How.

Within the first 8 seconds of a session, enough behavioral signal exists to predict which product categories have the highest conversion probability. This is what the data looks like.

Jennifer Nguyen