Insight, Impact, Innovation
HomeOur WorkCase StudiesPersonalized Targeting Boosted Redemption Revenue by 230%

Personalized Targeting Boosted Redemption Revenue by 230%

Download PDF

The Business Challenge

This client wanted to apply advanced analytics across 60 million households to derive shopper insights to ultimately increase sales. The insights would enable them to design more personalized and more relevant pricing, promotions, and product offerings for their customers.

The Solution

Fractal Analytics’ Customer Genomics® solution was customized to systematically learn the behavior of the retailer’s shoppers and products.

  • Hadoop, a set of Big Data algorithms for distributed storage and distributed processing, was deployed.
  • Hadoop was able to quickly and efficiently process this retailers transaction data from product sales, shopper data, social data and more over the last three years.
  • The solution defined personalized tags (called genome markers) for each customer across hundreds of different dimensions based on attitudes, behaviors, interests, and life stage events, as well as more traditional recency, frequency, and monetary metrics.

The Results

Redemption revenues jumped by 230% because of deploying more than 50 genome markers based on shopper attitudes, behaviors, interests, and lifestage events.

Example Customer Genome

Customer Genomics® develops more precise genomes over time as it learns with each transaction



Inferring shopper product usage patterns allowed us to find purchase patterns to inform offer structures tailored to different user groups. We were also able to identify the best performing offer combinations as well as those performing below par.


The Customer Genomics® deployment created a deep impact for this retailer:

  • Over 4,000 offers across 14 business units, powering more than 75 personalized campaigns each week.
  • A 20+% increase in revenue with smarter targeting with a decrease of 80% in default offer issuance.
  • An overall 230% increase in redemption revenues.


Automated machine-learning algorithms were applied to seeded client data to iterate until it developed razor precision, yielding over 50 genome markers for 60 million households in addition to genome markers for products and offers.