Lewis Perdue’s experience as a tech entrepreneur, wine importer, book and magazine publisher, Internet payments CEO, premium digital publications founder, author and consumer ingrained in him the vital importance of how accurate, “dead-on” recommendations drive repeat sales, build customer loyalty and increase profits. (More on Lew at Ideaworx.)
Lew’s hands-on work with, using, inventing and prototyping preference selection/recommendation engines — since his first system in 1994 — has shown him how every current system fails as often as it succeeds.
Shortly after building his first recommendation engine, Lew was one of the first to play with FireFly, the first collaborative filtering recommendation engine which invented the main paradigm (“people who bought this also bought this”) that still underpins today’s mainstream and episodically accurate efforts that attempt to make accurate recommendations for consumer purchases.
Experience with FireFly and subsequent generations of collaborative filtering “agents” and “bots” convinced Lew that a “top down,” external data-driven system would never be the system that consumers needed and wanted in order to buy with the confidence that they would like their purchases.
Perdue’s first recommendation system, Smart Taste, employed a profile matching system to help wine consumers. Building on the weaknesses of Smart Taste, Lew invented Savvy Taste which was coded by his son William and became the database core of a popular Facebook app of the same name. Savvy Taste’s greatest success was in uncovering the failures inherent in recommendation systems involving experts, social media, profile matching, and today’s big-data versions of collaborating filtering.
Perdue continues to develop improvements to recommendation systems.