General Merchandise Retailers2018Machine Learning (classification)Optimization / Operations ResearchB2C
Flos USA

Flos USA Lifts Checkout Conversion 125% with Full-Funnel A/B Optimization

By combining session recordings, heatmaps, and 30 targeted A/B tests over 17 months, the luxury lighting retailer turned a chronic conversion problem into an 18× ROI.

Checkout Conversion Lift125%
Program ROI18×
Listing Page Revenue Uplift86%
Duration17 mo.
5 min read

Background

Flos USA, a niche luxury lighting ecommerce site serving the US and Canada, struggled with low checkout conversion rates. The company lacked the internal resources and expertise to build a structured optimization program. In May 2017, it engaged VWO Services to manage end-to-end funnel optimization.

What Was Implemented

  • Behavioral research using VWO heatmaps, scrollmaps, and session recordings across all funnel stages
  • 30 A/B campaigns run over 17 months, each hypothesis grounded in quantitative and qualitative data
  • Homepage: navigation layout redesign to improve category discoverability
  • Product listing page: added CTA buttons to listing tiles
  • Product page: replaced dropdown color selector with visible color swatches
  • Cart page: stripped non-essential information, simplified layout
  • Full implementation handled by VWO Services team, minimizing internal resource requirement

Results

Over 17 months, the sequential funnel optimization program produced a 125% increase in Flos USA's checkout conversion rate and an 18× ROI from the overall CRO program. Stage-level gains included 86% more revenue from the product listing page, 19.35% lift in add-to-cart from the product page, and 36.97% higher add-to-cart rate from the cart page. The program is described by VWO as one client success story; all figures are vendor-reported for a single client.

Lessons

  • Full-funnel optimization outperforms checkout-only fixes: the largest gains came from listing and cart pages, not the checkout screen itself
  • Behavioral analytics (heatmaps, session recordings) are essential to generating testable hypotheses grounded in actual user behavior
  • Outsourcing test implementation to a specialist team allowed rapid iteration without internal headcount investment
  • Sequential, stage-by-stage testing compounds: each funnel stage improvement feeds volume into the next stage

Ready to implement AI in your commerce operations?

McFadyen Digital helps teams move from case study to live implementation.

Talk to an expert →