Recommendation Engine used to personalize CRM and e-commerce content drives 20% increase in revenue
Updated: May 29
The e-commerce platform for a national loyalty program had seen stagnant growth for a number of years, despite significantly increasing it’s offering from a breadth, depth and partnerships angle.
Given the increase in choice of available merchandise had not translated into greater revenue or site traffic there was clearly a marketing gap, but without a significant budget to drive awareness using paid channels, we needed to determine how to increase relevance and interest in the most relevant owned channels (email and the e-comm website itself).
After reviewing the previous marketing activities along with the e-comm and CRM platforms a couple of opportunities were identified where greater personalization could be introduced to driven interest and awareness.
Solution & Results
After reviewing different data science approaches that could drive the personalization, a SVD (Singular Value Decomposition) Recommendation Engine was selected as most relevant to the situation, given the volume of data and planned application. The Recommendation Engine would provide a list of the top 5 sub-categories or brands of product that each customer was most likely to be interested in, based on previous web, purchase and profile data. This was then used to personalize website content on their profile and basket checkout pages on the website, as well as in targeted email marketing, leveraging subject lines and dynamic content blocks in the emails.
The e-comm website saw a 5% increase in revenue in the subsequent month when the website changes were applied and an additional 10% once regular personalized emails were implemented to help drive incremental traffic to the site.