increase in RPM
Overview The client is an American global eCommerce marketplace connecting subscribers with local merchants by offering activities, travel, goods, and services in 15 countries. They enable real-time commerce across a range of categories including local businesses, travel destinations, consumer products, and live events.
Their mainstay is promotional discounts and coupons across segments and products.
By nature of the business and growing competition, the company’s growth strategy is tightly coupled with providing recommendations that are relevant to the location as well as the individual affinities.
This meant that the retailer needed to provide its customers with highly relevant product or service recommendations that not only aligned with their needs but also proximity requirements.
For e.g., recommending the right spa because the customer regularly purchases spa treatments is not good enough. Proximity to the spa is equally important for the customer to show interest and redeem the offer.
In order to drive personalized experiences across commerce site, web, and mobile app, the client deployed Recommend™ and Engage™—Algonomy’s personalized recommendations and content products—respectively. They leverage Algonomy’s profile service for segmentation and to understand user preferences & attributes in real-time. Merchandising rules are applied to support business needs while not compromising on the relevancy of recommendations.
With localization being a key requirement to drive individualized engagement, Algonomy provides a Geo-proximity feature for hyper localization of recommendation assets.
Shopper’s latitude-longitude information is juxtaposed with the latitude-longitude of the deals, which is determined using geo distribution of deals in a particular location.
This data is then leveraged to calculate the proximity of deals to the shopper in real-time and recommendations are re-sorted, so the most relevant deals closest to the shopper are presented on the top.
In other words, region-aware algorithms help in localization i.e., popular products in my region or area, and factors in proximity scoring to make relevant recommendations—products available near me. This is especially useful for deals from local service vendors.
The client was looking to compare the outcome of Algonomy’s configurable strategies with that of their generic strategies. To this end, Algonomy provided over 30 configurable strategies by leveraging different category seeds and user affinities.
These configurable strategies provide an easy way to generate relevant recommendations as backfills for Personalized recommendations on the Home page.
Using Top Selling strategy as a basis, recommendations were filtered based on an individual’s purchase history and affinity towards a specific merchant.
This resulted in a 250% increase in revenue per thousand impressions (RPM) and over 50% of attributable revenue on mobile app, mainly from Home Page.
Also, use of targeted Configurable Strategies on Commerce site and Mobile Web’s Deal Page, along with ‘Similar Items’ strategies, has driven over 80% of attributable sales.
User Affinities for Segments
The retailer explicitly asked their shoppers what their preferences were after they created an account, and wanted to market to those preferences on the site. They placed Algonomy tags on the preference page to send shoppers’ selections, thereby capturing preferences as user attributes in the User Profile Service.
With this, the retailer was able to set up specific segments to target certain shoppers.
Algonomy, a True Partner in Personalization
Algonomy drives strategic engagement through its Personalization consulting program and offers Personalization assessment, which covers audit and optimization recommendations.
Algonomy’s engagement with the client continues to expand with planned investments to improve user experience and engagement. This includes application of personalization across 12 international sites, additional placements on Browse, Search, and Cart pages, personalization based on IP location, and introducing replenishment algorithms.