Details: Recommendation Technology

Online-, shipping- and brick and mortar merchants, as well as media and service providers are increasingly confronted with clients who are well aware of their purchasing options: clients who compare options via the Internet, who consume a wide range of merchandise, and who increasingly allow themselves to be inspired by contributions from social media. Classic marketing is reaching the limits of its effectiveness for these clients.

The "informed consumer"...

  • ...uses search engines extensively to find product and pricing information,
  • ...purchases goods and services online,
  • ...increasingly expects variety and multi-media capability in connection with the purchase,
  • ...enjoys returning to the provider after a successful purchase experience.

Companies recognise opportunities and needs…

  • ...to present their total product range comprehensively and in an easily accessible manner, to confidently respond to purchase interest,
  • ...to recommend products based on interest, to create targeted incremental purchase interest.

Objectives:

  • Increased turnover
  • Increased customer retention time at the point of sale
  • Increased customer loyalty
  • Increased repeat visit rate

Challenges:

  • Long-Tail activation on par with top seller sales
  • Ensuring sufficient master data quality and a sufficient meta data foundation
  • Comprehensive tracking and evaluation of usage data, for instance for product evaluations, recommendation and claims
  • Targeted cross selling recommendations to increase turnover, while avoiding down buying effects
  • Giving consideration to environmental factors such as season, customer data or weather

The quality of the product range offered is decisive for high customer satisfaction and acceptance. Intelligent technology is needed to satisfy this requirement, especially for large product volumes, as well as for truly personalised presentations.

A success factor for the recommendation engine is the combined use of Pattern Science technologies, complementing "classic" machine learning processes for the analysis of customer behaviour with language processing capabilities and ontologies and/or semantic web.

This provides for the ability to develop customer recommendations with the help of content description analysis more precisely. Also, social media contributions on the Internet can be "crawled", giving you the ability to take advantage of user interests and comments when developing matching recommendations.

The targeted combination of these processes can be used to achieve optimum results in a variety of scenarios to best support the client in their purchase decision, and to provide for a truly personalised consumer experience.

<< Back


Twittern
LinkedIn
Bookmark on Delicous
Digg this
StumbleUpon