The Brand Context
LOFT is a casual women’s lifestyle brand for shoppers exploring across categories (tops, denim, dresses, lounge). LOFT customers expect fun, easy recommendations that flex from weekday to weekend.
The Challenge
LOFT wanted a recommender that captured browse first, feature-led behavior, surfacing playful prints, denim, knits, and seasonal capsules, while still solving issues with size fragmentation and customer preferences. The team hence leaned into Mapp Fashion’s apparel-specific models.
The experiment leading to Mapp Fashion
Two 2-way tests compared Mapp Fashion against 2 other players, including Salesforce Einstein. The initial proof of value began with Similar Items on PDP and expanding based on performance.
The Solution
- Discovery weighted recommendations that balance trend, context and physical attributes that genuinely reflect each customer’s preferences
- Algorithmic guardrails to respect promo and inventory/size depth, improving keep rate.
Results
- Mapp Fashion significantly outperformed the two other vendors, including Salesforce Einstein, on the key metrics of RPV and AOV.
Mapp Fashion’s apparel-only DNA was a key differentiator for us. Their models are built around apparel signals (fit, fabrication, occasion), not generic item similarity. We’re excited by the enriched product-data led, both in terms of what it delivered in the 3-way bake-off but also what it unlocks for us going forward.
Jordan Lustig
VP eCommerce @Ann Taylor
What’s Next
- Scale beyond PDP into homepage slots, cart cross-sell, and themed emails (e.g., “Weekday to Weekend,” “Cozy Layers”).
- Continue optimizing towards profitable growth by incorporating returns and customer preferences.
Discover Mapp Fashion ›