Retailers often balance personalization with trends and competitor benchmarks, but we wondered how much outside signals really reflect the reality of a retailer’s own audience. What if the patterns hiding in clickstream data told a different story altogether?
To explore this question, we analyzed over 10 million jeans purchases across three major retailers. Even in a single category, the results looked nothing alike. Each dot in the charts below represents a shade of denim, from light to dark washes. The bigger the bubble, the more customers bought it. As you’ll see, every retailer’s audience leaves a different fingerprint of demand.
What we found challenges the idea that apparel merchandising decisions should be driven primarily by trend reports or competitor playbooks. Trends and benchmarks can be useful inputs, but when they’re over-indexed, they risk producing false positives and surfacing products that may not actually resonate with your shoppers.
Your clickstream, on the other hand, can reveal which broader trends truly matter to your audience and filter out the noise. That’s why true personalization must be built from the behaviors on your own site at the foundation: over-weighting external signals risks obscuring the distinct patterns of your unique audience.
Brand 1: Depth Over Breadth
The first chart reveals a decisive pattern: customers at this retailer overwhelmingly prefer medium and darker washes. Most of the larger bubbles fall within the lower right quadrant.
For these shoppers, denim isn’t about chasing variety. It’s about doubling down on deep, saturated shades they already love. Expanding into lighter washes might make sense for competitors, but here the data makes it clear: this customer base consistently gravitates to dark and deep blue denim.
Brand 2: The Beauty of Choice
The second chart looks nothing like the first. Instead of a dominant cluster, demand is distributed pretty broadly across the grid. This retailer’s customers shop the full spectrum: light, mid-tone, and dark shades alike.
For this audience, variety isn’t a distraction — it’s the appeal. The challenge isn’t narrowing choice, but making exploration manageable so every shopper can efficiently find their “perfect pair.”
Brand 3: Small but Mighty Favorites
The third chart looks scattered at first glance: many small dots, with just a few outsized bubbles standing apart. But those bigger bubbles matter. This retailer has a handful of specific SKUs that inspire deep loyalty.
These aren’t the jeans everyone buys, but for the customers who do, they’re the exact right fit. The data highlights pockets of concentrated demand where devoted segments of the audience keep coming back for the same models.
Balancing Popularity with the Individual
The three charts above don’t just show clusters of best-sellers. They’re also full of tiny dots — the outliers. Each one represents shoppers whose preferences don’t match the majority.
A great personalization engine has to serve both: surfacing the products most customers want, while still recognizing and supporting the edge-case individuals. That balance is what makes the experience feel both highly relevant and personal.
Why Your Data Is the Only Data That Matters
Across three retailers, the same product category — jeans — produced three wildly different purchase patterns:
- Brand 1 thrives on depth
- Brand 2 thrives on breadth
- Brand 3 thrives on loyalty to specific SKUs
Trends and competitor benchmarks should influence merchandising, but they are external signals, and by definition, they’re averages. Personalization, by contrast, comes only from your customers’ own behavior. When external forces are overweighted, they can unintentionally flatten out what makes your audience unique.
Every retailer has its own fingerprint: a blend of products, shoppers, and geographies that can’t be reduced to industry averages. If these brands built their search experiences off competitor benchmarks or industry trends, they’d undermine their own strengths. But if they anchor personalization in their own clickstream data — both collective and individual — they can deliver experiences that capitalize on trends where appropriate while still feeling effortless, attractive, and uniquely theirs.
That’s why Constructor takes a different approach. We don’t optimize for what’s popular everywhere. We optimize for what matters to your customers, right now. Our search and discovery engine balances global demand signals with individual behaviors, ensuring that every shopper sees results that reflect both the wisdom of the crowd and the uniqueness of their own intent.
Because personalization isn’t about treating all products, or all customers, the same. It’s about letting your customer data lead — and using trends and competitor benchmarks only as context, not the compass.
Want to learn more? Download Beyond Relevance, the data-backed case for attractiveness as the new standard for ecommerce search performance.
