Case study — Retail · E-commerce
A recommendation engine and unified commerce platform across eight storefronts for a multi-brand retail group.
Client
Multi-brand retail group
Scope
E-commerce · Personalization · Engineering
Timeline
5 months
Footprint
8 storefronts · 1.2M monthly shoppers
01 — Challenge
Eight storefronts, one generic homepage — recommendations that ignored what shoppers actually did.
Every brand ran its own inventory and merchandising logic, so nothing learned across the group even when shoppers overlapped.
Recommendations were manually curated and rarely updated, missing the browsing and purchase signal already being collected.
02 — Approach
PHASE 01
Research
Audited existing inventory systems and shopper behavior across all eight storefronts.
PHASE 02
Design
Unified merchandising rules that could flex per brand without a rebuild per brand.
PHASE 03
Development
Built the recommendation engine and unified inventory/OMS as one integrated layer.
PHASE 04
Rollout
A/B tested against the existing homepage before full rollout across all storefronts.
03 — Solution
Built for the workflow
01
Recommendation engine
Product suggestions driven by real browsing and purchase signal, not manual curation.
02
Unified inventory & OMS
One inventory source of truth across all eight storefronts.
03
Personalized merchandising
Homepage and category pages adapt per shopper without per-brand engineering work.
04
Growth marketing integration
Recommendation data feeds directly into retargeting and email campaigns.
StackNext.jsTypeScriptPythonRedisPostgreSQLShopifyGCPSegment
04 — Outcome
22%
Lift in conversion
1.2M
Monthly shoppers served
8
Storefronts unified
3.1x
Return on ad spend
“The recommendations finally reflect what people are actually browsing, not what a merchandiser guessed last quarter.”
VP eCommerce — Multi-brand retail group