Fascinating new research from the Wharton School and CMU on today's product recommendation systems:
- They don't really promote the long tail
- Long tail recommendations are more impactful than recommendations of popular items
- They are more effective for "sensory" products (eg jewelry) than "functional" products like appliances.
See details at http://knowledge.wharton.upenn.edu/article/recommended-for-you-how-well-does-personalized-marketing-work/
Cognilyze's psychology based product recommendation engine EXACTLY answers these gaps in today's product recommendation. Our recommendations are popularity-agnostic, recommending the best-matching product for each person, from the most popular to the narrowest niche. Our system is also highly focused on the motivations underlying functional products as much as on the personal preferences underlying sensory products.