Client
Mid-size D2C apparel brand in India with growing paid traffic but low conversion from product listings.
Challenge
Low add-to-cart rate and generic catalog experience caused high bounce from PLP and low PDP→checkout CTR.
- ▸Add-to-cart rate at 2.3%
- ▸Generic PLP/PDP with no personalization
- ▸No on-site search re-ranking by intent
- ▸Limited experimentation capability
Solution
Implemented AI-driven recommendations and personalization across PLP, PDP, and cart with A/B testing guardrails.
- ▸Session + history based product recommendations (home, PLP, PDP, cart)
- ▸Segments for new vs returning vs high-intent users
- ▸On-site search re-ranking based on intent and margin
- ▸Experiment framework with feature flags
Tech Stack
- ▸React/Next.js, Node.js
- ▸Python services, Redis, PostgreSQL
- ▸GA4, Segment, BigQuery for analytics
Implementation
- ▸Week 1–2: Data audit, tracking, ETL to BigQuery
- ▸Week 3–4: Recs MVP + PDP/PLP containers; flags
- ▸Week 5–6: Search re-rank, cart upsell, experiments
Results
- ▸+31% add-to-cart rate (2.3% → 3.0%)
- ▸+18% PDP→checkout CTR
- ▸+12% AOV via bundles/upsells
- ▸-9% PLP bounce
Business Impact
Improved ROAS from paid campaigns and reduced merchandising overhead due to automated recommendations.