Identity Shop Case Study
Start a project All work

Case Study 2026 AI Commerce · Personalisation

IDENTITY

Dom
Conversion lift
3.2x 
Identity dimension
10 
Products mapped
98 

Overview

Identity Shop is an AI-powered product recommendation engine that maps shoppers across 10 psychographic dimensions — risk tolerance, aesthetic preference, values, lifestyle — and serves hyper-personalised product recommendations. Conversion rates are 3.2x the category average.

Traditional e-commerce treats every shopper the same. Product recommendations are based on what other people bought, not on who the individual is. This works for commodity products but fails for identity-driven purchases where personal values, aesthetics, and lifestyle matter. The result: low conversion, high return rates, and a generic shopping experience.

What was built

  • ✓Psychographic profiling system mapping 10 identity dimensions
  • ✓Claude AI integration for real-time personality analysis
  • ✓Product matching algorithm across 98 products
  • ✓Conversational shopping interface
  • ✓Identity visualisation dashboard showing shopper profile
  • ✓A/B testing framework for recommendation strategies
  • ✓Analytics tracking conversion by psychographic segment

Approach & Tools

Key decisions

  • Profile the person, not the purchase history
    Collaborative filtering recommends what people like you bought. For identity-driven purchases, where values, aesthetics, and lifestyle decide the sale, that misses. Identity Shop maps the shopper across ten psychographic dimensions and matches on who they are, not on a lookalike’s cart.
  • Let Claude build the profile in conversation
    Behavioural personalisation needs months of clickstream before it works. Using Claude to read a short conversation means the profile is rich from the first session, no cold-start, no waiting for data to accumulate.
  • Prove the lift, don’t assume it
    Personalisation is easy to claim and hard to measure. An A/B framework and per-segment analytics were built in from the start, which is how the 3.2x conversion lift over the category is a measured number and not a hope.

Inside look

Next case Three Month Rule
  • News
Book a call Start a Project