As User Sophistication Evolves, the Need for Recommender Systems Increases
As corporate and retail user sophistication continues to evolve in tandem with ever growing market pressures and often groundbreaking shifts in many verticals, (passing through increasingly shortened and pronounced economic cycles), companies have declared the capacity to be able to react and to anticipate a mission critical necessity and not a mere nice-to-have for an elusive future state.
As pointed out by the authors of the paper “Session-Based Item Recommendation in e-Commerce: on short-term intents, reminders, trends and discounts” (Dietmar Jannach, 2017), in the world of e-retail, advanced features such as recommender systems are of utmost importance when it comes to purchase suggestions that are on the one hand, based upon historical data analytics but also, on the other hand, on cross-referenced, virtual interest footprints the consumer can bring to the table. These and other functionalities guide and improve the customer journey significantly but also require a sound e-commerce solution architecture that not only allows, but in fact empowers agile market responses. A sluggish and slow multi-channel front end won’t entice the busy B2B end user nor will it win over the short attention-spanned mobile shopper.
In their experience at Coreshop Solutions (https://www.coreshopsolutions.com) and NSI Solutions (https://www.nsisoluciones.com/eng/), at least two underlying components have to liaise in harmony for an e-commerce platform to optimally leverage the wealth of hundreds of thousands, if not millions of different data points, enabling the final sale: Product Inventory Management (PIM) and Business Process Management (BPM). Over are the days where it would be recommendable to contain all the product and services information, including the multiple layers of derivative data, directly on the e-commerce site. The strength and advantages of PIM to the end users come into play especially in conjunction with its Master Data Management capabilities. Vast amounts of data can be efficiently stored and, more importantly, then be used from several front facing channels at fast speeds, in parallel. When it comes to complying with the often very complex business rules, workflows and approval matrixes that typically accompany a modern e-commerce, end-to-end solution, BPM appears to provide a perfect component to the overall engine: agile and intuitive process automation. Scenarios like data ingestion management, special approval processes for customer requests, customer onboarding and scoring will become a many-fold, easier to implement, integrate and to be managed by the nontechnical staff.
This reasoning, naturally, extends also to the adjacent process automation category of Robotic Process Automation (RPA) under which microservice can be automated. A common scenario for RPA in the e-commerce realm turns out to be within the field of data reconciliation, cleansing as well as regression test automation.