Data and AI, Machine Learning
GoDataDriven | Part of Xebia
The customer is a top-three internet retailer in the Netherlands that serves 1.7 million customers per year and records a total of 145 million site visits.
Additional customer interactions and engagement for the retailer take the form of online reviews, calls to its contact center, and email communications, where email communications open and clickthrough rates are measured.
Generating Greater Value by Personalizing Shoppers’ Experience
Personalized customer journeys can deliver sustained business value for internet retailers. As an early adopter of this transactional mapping process, the e-tailer found success in creating static profiles for groups of similar customers. But as technology and shopping behavior have evolved, the algorithms that power retail platforms can now be individually tailored to each customer’s journey. The more finely tuned these algorithms, the better they will be able to respond to individual customer interactions such as increased search and click-throughs, and then influence purchase decisions. For the e-tailer to maintain its leadership and retain customers, it was essential to develop a new platform that collects and analyzes this customer data in real-time and at a low cost.
Building a single analytics platform using a three-plank approach
The retailer wanted to personalize customer journeys using smart algorithms on a single platform, soXebiacreated a three-pronged strategy: technology upgrades, new methods of working, and skilled human resources. The first step was consolidating shoppers’ data from assorted existing legacy systems, many of which had been developed in-house. A new platform would then need to be built to make sure that all data from the underlying systems was made available in the distributed file system Hadoop. At the same time, new ways of working were required to speed up product development. Finally, the retailer would need to recruit and train new staff, pivoting from business intelligence professionals to data scientists.
A Hybrid Solution of Agile, Scrum, and DevOps
Open-source solutions were the first software option for the new platform. For example, to capture clickstreams from the website, the retailer opted to use Divolte. Where open-source applications were not available, programs were developed in-house. Once the platform was built, the resulting smart algorithms offered an insight into customer needs based on data, such as order history and real-time site interactions. The organization now uses Agile, Scrum, and DevOps to develop its software. Teams act as product owners and are responsible for tracking and furthering progress. This culture shift started at the IT department and has since spread to other departments, underpinning the retailer’s new approach to customer journeys.