Product • 26.08.2008
prudsys RECOMMENDATION ENGINE
The prudsys RE is a recommendation engine that submits appropriate recommendations for various products and users in real time. Thanks to its intelligent methods in the real time context it goes even further: it enables high quality simulations, optimisations and reports in the field of range design and pricing.
The prudsys RE is intended for retailers and is meant primarily to bring about yield increase via customer interaction systems like online shops, in-store POS (kiosks, tills, scales, PSA, PDA, etc.) and call centres. The recommendation engine is multichannel-capable: analyses can be run combined over several channels and vice versa, recommendations can be submitted via several channels.
RE scenarios in online business
Product recommendation
Product recommendations for individual products are generated, in particular on product detail pages.
Search recommendation
Appropriate recommendations are provided for a given search term. In this case we are dealing with the viewed or purchased products combined most often with the search term. The search recommendation is transaction-based and thus a valuable complement to the conventional, content-based search.
Market research
Product recommendations and prices are manually provided and tested in the recommendation engine. The integrated reporting of the prudsys RE then uses control groups to evaluate how successful the measures were. This generates new approaches for verifying the range and pricing policy.
Consultants
Virtual consultants and avatars are controlled via the prudsys RE reinforcement learning in order to optimise their consulting performance.
Personal recommendation
Depending on previous transactions (clicks, purchases, etc.), personal recommendations are displayed on the shopping basket or "My Shop" page, for example. The real time scoring module also allows the use of user master data for highly personalised recommendations.
Newsletter
Personalised newsletters are created and are thus much more efficient. This is made possible by using the mechanisms for personal recommendation in combination with special features in the newsletter module.
Campaign management
The prudsys RE sends out impulses for multichannel campaign management via the relationship management engine module. In so doing it covers complex action chains in addition to follow-up mails with special recommendations following cancelled shopping basket processes.
Price optimisation
The major topic of the future! Based on clicks and purchases, the dynamic price optimisation module automatically varies the prices of products or categories. Special promotions and bargain shops are ideal as an entry point.
Reporting
The prudsys RE allows the creation of continuous, comprehensive statistics (especially when using control groups) using special log charts. These can be used via the reporting module for interactive OLAP analyses and automatic reports.
RE usage scenarios in stationary business
Product recommendation
In this case, product recommendations for individual products are generated. They are inputted by hand, scanned in or automatically detected (e.g terminals, scales, PSAs, PDAs, etc.).
Arrangement
The prudsys RE simplifies meaningful arrangement of products on shelves and on topic tables. In so doing the recommendation engine interacts with special PDAs of the scheduler, which can then interactively optimise the range.
Market research
Product recommendations and prices can be manually generated and tested using interactive devices in the recommendation engine. The reporting integrated into the prudsys RE then evaluates the success of the measures using control groups and verifies approaches to range and pricing policies.
Consultants
Virtual consultants and avatars are controlled via the prudsys RE reinforcement learning in order to optimise their consulting performance (used predominantly via terminals).
Personal recommendation
Depending on previous transactions (inputs, scans, purchases etc.) personal recommendations are put together, which can be issued, for example, using tills with checkout couponing functions. The real time scoring module also allows the use of user master data for highly personalised recommendations.
Campaign management
The prudsys RE sends out impulses for multichannel campaign management via the external relationship management engine module and comprises complex action chains in addition to post-purchase loyalty letters.
Price optimisation
The major topic of the future! Based on sales over periodic periods of time as well as product recalls if necessary (entries, scans, etc.) the dynamic price optimisation module automatically varies the prices of products and categories. Entry via promotions, especially couponing.
Reporting
The prudsys RE allows the creation of continuous, comprehensive statistics (especially when using control groups) using special log charts. These can be used via the reporting module for interactive OLAP analyses and automatic reports.
