According to García-Noriega, not all customer requests are specifically for big data; most of the challenges and demands that usually apply for big data apply for retail analytical needs too. “Customers often assume that by acquiring one single product, they have access to all big data benefits. Also, the assumption that with big data, business users become analytical thinkers and immediately start churning business insights is vague,” explains García-Noriega. “We offer customers a packaged offering that implements and monitors a recurrent analytical process for their customers.” This packaged offering is in reality SaaS scheme that covers all the key components necessary to fulfill the big data promise.
The company developed Mermaid, a big data portfolio for retail analytics giving customer the freedom to choose from predefined big data architecture options. It packages all the required data warehouse layouts, data loading and movement processes, KPIs, and optimization techniques to optimally run and quickly implement the complex retail analytical foundation. Once the foundation is in place, customers can immediately put to use the many Data Mining libraries already incorporated into the Mermaid modules to have a deeper understanding and better answer common retail problems including proper product segmentation and allocation, excess inventory on stores or price increases affecting demand volume.
We offer customers a packaged offering that implements and monitors a recurrent analytical process for customers
The company has also developed their own BIAS (business intelligence as a service) methodology that views the analytics initiative in customers as an ongoing process that needs constant data monitoring, dashboard improvements, and user support. On the other hand, RITTER’s Elephant application handles raw weblog data offering retailers a true big data solution to gather data from their e-commerce and social media activities.
In an instance, one of RITTER’s retail clients had millions in overstocked inventory but could not accurately pinpoint in which products and stores the inventory was located and the exact reason the inventory had accumulated. With a deadline in mind, RITTER created a decision tree that grouped products in two segments and then using a combined DeltaMaster, SQL Server, and Netezza architecture, they assisted the client to locate and remove more than a third of the cost from this overstocked inventory in the first month. “Afterwards, the solution was automated and nowadays, the customer quantifies through a visual dashboard the overstocked inventory,” notes García-Noriega.
RITTER is keen on R&D and invests regularly in their DataLab. The lab constantly researches data mining and machine learning techniques to solve current retail problems. “We are expanding our big data retail offering to the retailers supply chain and currently developing the Bullwhip line. It is a modular solution geared to measure variability in the supply chain and discovers patterns and inconsistencies,” concludes García-Noriega.