Self-learning and multi-channel predictive recommendation


Data collection

Aggregation of various data: behavioral data (visits, clicks, add-to-cart), historical data, user data (social, segmentation)

Graph creation

Generation of direct and indirect relationships between users and products


Analysis of relations, according to multiple criteria. Then calculating a puchase probability computation


Real time recommendations display in accordance with each profile

Multi-channel and comprehensive solution

« Conversion »
Mobile and website recommendations.

Display 100% personalized recommendations from the 1st visit, on every page from the home to the shopping cart page, tailored to your ergonomics.

« Loyalty »
Email recommendations.

Customize your newsletter, retarget abandoned carts, add commercial bounce down in your transactional emails and detect customers who desire a specific product.

« Acquisition »
Cross-site recommendations

Display cross product/content recommendations  from third party websites to create traffic synergies between your sites galaxy.

Automated and self-learning solution

Business rules management

Consideration of specific business rules : inventory management, margin management or management of product grouping constraints.

Recommendations automation

No manual settings required at any time.


improvement of the algorithm according to successes and failures observed, thanks to a machine learning system.

Business partner approach


Business partner approach

  • Analysis of purchase history and pre-launch modeling for rapid effectiveness of the recommendations.
  • Regular algorithm optimization to improve performance recommendations.

Performance-based model

  • Low cost setup to quickly optimize results.
  • Fees on the sales generated by the recommendations.