Click here to return to the Bright Angles website

Sunday, 12 August 2012

Learning from Amazon's recommendations system

Today we share an excellent article by J P Mangalidan at Fortune Tech, detailing out Amazon's method of providing recommendations to users.

Anyone who has used Amazon would have experienced the persistence  of the e-tailer in offering recommendations. Amazon integrates the user's purchase behaviour, browsing and reviews read to serve up customised recommendations through email and in panels of products that you see while browsing the site.

Given that Amazon is a huge store with hundreds of departments and thousands of brands, it makes sense that the company has to create a mechanism for people to discover products that are relevant to them. Providing recommendations is a way to increase the ARPU, to create loyalty and repeat purchase and to bond with customers by showing that the site understands them and their requirements.

We have nearly always found Amazon's recommendations to be accurate and to reflect our interests. In the rare cases where they have been off target, they provide users with the flexibility to edit and change the settings. The ecommerce giant's keenness to learn what users like is impressive and definitely helps them to improve sales and revenues.