The professional solution for recommendations provided by YOOCHOOSE for the Nestlé Schöller B2B online shop offers a service that is adapted to the B2B environment. The perennial challenge in a B2C environment consists in animating the customer to make a one-time purchase of a product, preferably a product of relatively high value. The B2B environment on the other hand is characterized by distinct features: Business customers are usually tied to a few specific suppliers in the long run and regularly purchase large quantities of a very static basket of products taken from the entire product range. The Recommendation Engine has to take these differences into account in order to consistently increase customer lifetime value and customer retention in a B2B environment. This has been proven to work with the integration of the YOOCHOOSE Recommendation Engine in the Nestlé Schöller online shop.
The project partner
The NESTLÉ SCHÖLLER GmbH is an expert supplier in the areas of ice cream, frozen baked goods and other frozen meals in the out-of-house market. The SCHÖLLER DIRECT B2B-Online-Shop only serves commercial buyers in the areas of full service, quick service, leisure time, catering, commerce and logistics. SCHÖLLER considers itself to be not just a delivery service, but a partner who fulfills the highest expectations. They manage to delight customers over and over again thanks to unique product concepts, highest quality standards, individualized service and personal support.
Challenge and goal
Working together with the integration partner ETECTURE, we tried to achieve several goals by integrating recommendations into the online shop. On one hand, recommendations are supposed to facilitate use of the shop and the process of finding products, thereby making the shop more attractive for customers. On the other hand, recommendations are supposed to lead to noticeable cross-sale and up-sale effects by making recommendations for products that the customer was not previously familiar with and didn’t search for. Finally, the recommendation engine also serves to analyze buying behavior and provides marketing measures such as personalized newsletters and automatically generated shopping lists.
During the first phase of the project it was decided in which parts of the shop recommendations should be used and what types of recommendations and filter settings should be deployed in the chosen locations. Furthermore, user tracking was integrated into the shop in order to measure activities such as clicks, shopping baskets and buying actions. Following that, our partner ETECTURE began to implement designated areas in which the recommendations were put to use. A special feature of the YOOCHOOSE Recommender Engine is the fact that recommendations can not only be delivered as finished, rendered HTML code but also as JSON objects called upon using a REST API. This allows the shopping system to alter any conditions or remaining product quantities displayed in the recommendations even as the session is still ongoing.In order to prepare the product catalogue and product attributes to be suitable for recommendations, a nightly reconciliation was established for the product catalogue. This was also completed using YOOCHOOSE’s flexible REST API, which provides several methods for data import and reconciliation as a standard feature. Following concluding tests and delivery to the customers, the shop and the newly implemented features went online just 5 weeks after the project start. In order to shorten the cold start phase and to better conform to the typical B2B buying rhythm, which is marked by regularity and a very high share of repeat product purchases, an import of the entire customer base’s shopping history going back a year including purchases made by fax and telephone was also conducted. This meant that the YOOCHOOSE Recommendation Engine was capable of making high-quality recommendations from the start, without having to first track and determine the clicking and buying behaviors of customers over a prolonged period of time.
Using the YOOCHOOSE Recommendation Engine, Nestlé Schöller has, within a remarkably short time period, deployed a tool that allows a targeted use of cross-selling and up-selling recommendations within the online shop. Thanks to the YOOCHOOSE Recommendation Engine’s flexible and comprehensive API, special characteristics of the B2B business could also be considered without investing a lot of time into elaborate adjustments to the standard solution. Aside from having low implementation costs, the solution also reduces Nestlé Schöller’s future costs for upkeep and maintenance to a minimum. This proves once again that software as a service does not necessarily have to be inflexible and that it can tackle a wide array of different of needs if a good API is available and made use of.
Project-Partner: ETECTURE - DIGITAL ARCHITECTS