Personalization Experiment for a leading Indian Retail Brand
Validating the measurable revenue impact of personalizing home page experiences before scaling.
Executive summary
The brand aims to evaluate the tangible impact of personalizing homepage experiences before scaling personalization across the entire app.
The objective is to test whether segment-based personalization of a single homepage component can deliver measurable uplift in engagement and revenue.
If clear uplift is observed, the data will be used to justify scaling personalization across all key pages, more audience clusters and components.
Experiment overview
Primary goal
Measure the behavioral and revenue impact of introducing personalized content in one strategic homepage component.
What Was Tested
To do this, we set up an A/B experiment targeting the 6th banner on the homepage, testing personalized experiences for defined customer segments vs. a non-personalized control group.
Customer Base
users landing on Homepage, 10% of sessions
Random Split
50:50
Control group
Generic 6th banner
Test group
Cohort-based personalized
Methodology
Two groups were created.
Test group
Users are assigned to one of the personalized experiences based on their cohort.
Control group
A random mix of all audiences, having a non-personalized version of the 6th banner.
Segmentation Framework
Two behavioral dimensions:
Premium vs. Mass purchasing behavior
Based on AOV and share of purchases from premium brands.
Preferred shopping category
Men, Women.
Process & Setup
Objective
Measure the revenue impact of introducing personalized content in one strategic homepage component.
Customer segments
Customers are randomly assigned to control or test groups.
Setup
All users are first exposed to a generic 6th banner – then shown the banner specific to their segment. Control users of cohort see the same non-personalized banner. Where as test users of mapped cohort see dynamic banners based on mapping.
Experiment duration
14 days
Cohorts Used
Four cohorts were defined based on behavioral dimensions:
Premium men
- •AOV > ₹2,000
- •> 50% purchases from premium brands
- •Men's category share > 60%
Mass men
- •AOV < ₹2,000
- •or < 50% purchases from premium brands
- •Men's category share > 60%
Premium women
- •AOV > ₹2,000
- •> 50% purchases from premium brands
- •Women's category share > 60%
Mass women
- •AOV < ₹2,000
- •or < 50% purchases from premium brands
- •Women's category share > 60%
*List of premium brands was provided by the business team.
KPIs Measured
We tracked four critical metrics to measure the impact of personalization on engagement and revenue.
CTR (Click-Through Rate)
Percentage of users who clicked on product/link in the message.
Session conversion rate
Users who completed purchases within the session.
Revenue Uplift
Average transaction value per exposed user.
Incremental Revenue
Total revenue uplift generated by personalization.
Experiment results
The personalized homepage component demonstrated substantial improvements across all key engagement and revenue metrics.
Conversion rate
Revenue per user
Total revenue uplift
Key Finding
Personalizing a single homepage component delivered a +105% uplift in CTR and 1.72x incremental revenue from that component.
Brands and categories were selected based on audience segment preferences, not generic recommendations.
This led to higher CTR and improved conversion rates.
By exposing premium users to premium brands, we saw:
Higher AOV, Higher ASP, No change in return/pass rate
Overall:
Strongly validates the business case for scaling personalization across high-traffic surfaces.
Impact
From a single component in landing page campaign with cohort based personalization.
