InfiniteInsight® is driving personalized product recommendations to millions of unique website visitors every day.
Business Intelligence Director
Allegro Boosts Click-through Rates by 500% with InfiniteInsight®
- The Allegro Group, the biggest non-eBay marketplace platform worldwide, wanted to offer personalized product recommendations to its 20+ million distinct website visitors.
- Allegro needed to increase the income and gross merchandise volume (GMV) of its auction and e-Commerce websites in 11 Eastern European countries through cross-selling and up-selling to visitors.
- Low click-through (CTR) and conversion rates of generalized “bestseller” rules.
- With 15+ million listed products for sale, 100 million requests per day, and 500 million daily page views, prior attempts to deploy online recommendations were futile based on the large volumes of data and daily transactions.
- Standardized on KXEN’s InfiniteInsight® for personalized product recommendations and propensity models for newsletters.
- After having benchmarked more than 5 vendor solutions, Allegro chose KXEN’s InfiniteInsight® Social for its ability to analyze huge volumes of online data, its ease of use, and its flexibility to adapt to the company’s unique website requirements.
- Deployed InfiniteInsight® Recommendation as the foundation of the company’s website recommendation engine, using social network analysis to link products and create weightings based on visitor clickpaths, items placed in shopping carts, and purchase transactions.
- Integrated the recommendation engine with the company’s Oracle Exadata data warehouse and plan to integrate the propensity models with Neolane’s campaign management system in the future.
- Over 80 million personalized product recommendations to Web pages per day.
- Increased page views per visit by 30% and click-through rate on recommended products by 500% (from 0.54% to 3.09%) as compared to “bestseller” rules when visitors are reviewing their shopping basket.
- Conversion rates increased 40x with personalized product recommendations as compared to displaying only bestsellers.
- Able to serve product recommendations to website in fewer than 200 milliseconds with recommendation rules which include scoring by InfiniteInsight®.
- Build over 160, 000 rules every week from 118,785 product links (shoppers that “watched” similar products) and 43,074 purchase links (shoppers that purchased similar products).
- Automated the end-to-end recommendation process, from data preparation to building product recommendations, and applying models and scoring rules in production to their website.