LAS VEGAS -- Splitting hairs between what business intelligence (BI) and artificial intelligence (AI) mean could be a frivolous game of semantics, but for retailers hoping to unlock the transactional secrets stored in their point-of-sale (POS) registers, understanding the differences could prove crucial, according to a technology session speaker at the 2018 NACS Show in Las Vegas.
Addressing attendees at the workshop “TechEdge: Using AI to Improve Your BI,” Nabeel Azar, COO of Verisk Retail, Bloomington, Minn., said the main difference lies in what each technology focuses on. Essentially, BI tools look toward the past, assessing the information derived from transactions done over a period of time. AI tends to be more forward looking, predictive of what could happen.
“To understand AI, you need to look at BI as a way to deliver insights to you,” Azar said. “AI and BI pair together.”
With AI, concepts such as “machine learning” and “deep learning” are ways to get computers to solve future problems, Azar said.
Such problems include the following:
- Knowing the customer. Topics here include identity management, understanding how customers navigate the store and how retailers can best market to them.
- Monitoring. This area would encompass things like loss prevention.
- Supply chain and inventory management. Here a retailer would focus on product availability, price optimization and finding patterns in data.
- Customer support. In this area, automated communications for common transactions could save time and labor costs.
Gearing up to use BI and AI solutions means understanding the company’s automated “ecosystem” (or the machines exist within the company and what information they store); the problems needing to be solved; and the “modeling” of data that needs to happen to reap actionable insights, he said.
Ultimately, a company needs to have a “data-driven culture” in order to benefit from BI and AI technologies, Azar said. To get to that place, Azar suggested companies start with a tangible, commonly shared problem. Using data to actually solve that problem will be a motivating exercise.
“Find a problem that people want to find the answer to,” he said. “The company has to want to be data driven.”
Nabeel Azar, Verisk Retail. Photograph by CSP Staff