I just returned from NRF 2017--where the retail information technology community converged to chart the course for the year ahead and beyond. I went to test my point of view that applying artificial intelligence in retail isn't science fiction nor needs to be a science project. I came away more convinced than ever that it's time for action. Here's a snapshot of what I found.
Whether you call it artificial intelligence (AI) or cognitive computer, there was plenty of evidence at NRF 2017 that AI is the next best action for retailers looking to improve performance across marketing and customer engagement, merchandising and assortment, omni-channel commerce, and supply chain and operations. Looking at that expansive remit might cause a dilution of effort and blurring of focus—threatening two important prerequisites for success. While the opportunity calls for caution, retailers should guard against trepidation just as much as against unbridled enthusiasm.
We saw a number of credible AI offerings this week at the Javits Center. They're the ones that met—or at least approach meeting our URLIE hurdles of AI credibility. To one degree or another they demonstrated capabilities to understand varied content, reason through disparate data to draw conclusions and recommendations, learn as they go, interact with knowledge workers and customers, and explain themselves. The last two attributes are just as critical as the rest, if not more so. They're necessary for building line of business and consumer confidence. The pace of AI adoption and benefit depend on trust.
Of course, there's Watson, and IBM continues to deepen its capabilities and extend them across its commerce, marketing, and supply chain applications. Retailers pursuing AI should understand that "Watson" means three things—a core IBM business unit central to the company's future, a set of API services to power cognitive capabilities in its ISV ecosystem, and a badge of distinction earned by IBM applications that stand up to the URLIE test described above. IBM's high level messaging focuses on the first of these. Retailers need to focus their attention on the latter two.
But the AI line up at NRF hardly stops with IBM. Other companies bringing credible AI systems to market come from varied quarters. Wipro and Infosys are notables in the SI community with their Holmes and Mana AI platforms. Both demonstrated AI in customer engagement and insight and in support of knowledge worker productivity and efficacy.
Beyond the SI community we saw AI coming from innovators in marketing where Adobe, AgilOne, and Optimove stand out and omni-channel commerce where Boomerang Commerce came to the fore. While still gaining its footing among retail "ERP" companies, INFOR can claim a place among AI leaders in retail with its Predictix and Dynamic Sciences Lab assets. At the other end of the spectrum Kore caught our fancy as an AI-based conversational commerce bot.
Blue Yonder and Sentient Technologies are AI "pure plays" that have trained their sights to apply AI in focused use cases. Both deserve a special callout.
With the core of its science bench coming from CERN, the European Organization for Nuclear Research, Blue Yonder is riding high helping retailers in various segments in forecasting, replenishment, promotional buying, pricing, and markdown optimization.
Sentient Technologies, with a leadership and IP heritage tracing back to the developers of Apple's Siri intelligent assistant, is applying AI in two dimensions. It's already in production at leading retailers to curate visual search and discovery, merchandising personalization, and recommendations in footwear, apparel, and home furnishings. It's also bringing evolutionary algorithms to AB testing to rapidly optimize landing pages and customer journey funnels. It's handle optimizations involving as few as 360 combinations of text, images, and layouts to as many as 380,000 combinations.
Finally, let's not forget SAS—the world's largest privately-held analytics company. That stalwart has plenty of AI and machine learning capabilities deployed in various industries and use cases. SAS clearly intends to bring these capabilities into at least two retail applications, customer insights and assortment optimization. SAS Viya, its open platform for analytics innovation, will be the foundation for deployment of AI micro-services in these areas.
For more information, click here to register for the IDC Retail Insights 2017 Web Conference Series.