At NRF, fun and interactive AI exhibits dominated the scene. But glitzy as they were, they don't represent the full scope of AI-powered systems, which often work behind the scenes.
At NRF 2020, held at the Javits Center in Manhattan this past Sunday through Tuesday, the lines between exhibitors and attendees blurred to an unusual extent, as attendees strolling the expo floors were treated to a number of displays that bounced their own likenesses back at them. The culprits? AI companies like AnyVision and Ultinous, which lured scores of attendees with video monitors displaying facial recognition and analytics in real-time. Impira’s booth happened to be stationed kitty-corner to the menhir-like walls of AnyVision’s kiosk, and so we saw the steady stream of foot traffic in double, with AnyVision’s cameras registering the face of each and every passerby.
AnyVision specializes in facial and object recognition for mass crowd events, but its computer vision algorithms are far more discerning than a generic security cam. AnyVision will not only alert store managers when a high-spending repeat customer moseys through the door, but will also use “Gaze Estimation” to log that customer stopping to stare longingly at an item just barely out of budget—a cue to the sales clerk down the aisle.
Walking by the Ultinous booth, I glimpsed my own face framed by a tracking rectangle. Upon closer inspection, I noticed an integer on the rectangle’s border, which flitted between 18 and 30 depending on how I tilted my head—the algorithm’s best guess as to my age. Ultinous uses algorithms such as this one to provide retailers with greater demographic insight into clients’ browsing habits. If you mount enough cameras in your store, you can quickly get a sense of the typical walking routes of 25-year-old visitors versus 45-year-olds, then feed this data to your merchandisers to let them optimize accordingly.
In short, the event’s flashiest AI expositions were decidedly customer-oriented, aimed at persuading retail execs of the importance of knowing more and more about their customers’ habits. Yet there were also AI companies oriented in another direction: toward helping retailers improve their own internal processes and data flows—an equally important venture, though one less likely to elicit awe on the expo floor.
One example is CB4, an AI company that uses transactional data to track anomalies in retail performance. For example, if a given shoe store in a particular geography reports underwhelming numbers for a product selling well at neighboring stores, CB4 pings the store manager and suggests possible causes of the underperformance, along with actionable solutions.
At Impira, our technology serves a similar purpose, only on the digital plane. Just as a store manager not leveraging CB4 might fail to notice the glaring flaws of a particular product display, a brand manager using an old-school PIM system might fail to notice that a particular product is losing out on Amazon or Walmart.com due to low-quality visuals or incomplete and faulty product data. Because pre-AI systems have no way of autonomously aggregating or enriching data, companies reliant on these systems are by definition reliant on manual data entry. For companies with thousands of products to manage (and numerous images and creatives associated with each of those products), AI-based systems like Impira’s deliver a significant competitive advantage by quality-controlling digital displays to the utmost degree.
E-commerce companies invest heavily in analytics tools that are effectively the digital equivalent of an AnyVision or Ultinous, providing insight into online browsing habits as part of the ongoing thrust toward personalization. But personalization demands more than a well-curated CRM; understanding customer tastes is only half the battle. E-commerce companies also need tools that expedite the creation and curation of the digital displays and experiences necessary to deliver customers exactly the right content at the right moment. Tools of this nature are by no means eye-catching, but they power the critical behind-the-scenes operations that allow customer-oriented insights to produce real revenue impact.
Knowing that Joe Schmo is neck-deep in a Hawaiian shirt phase is one thing. Having the right ad(s) handy to persuade him to pick your Hawaiian shirt over a competitor’s is another. Many e-commerce teams today struggle to act on the insights they gather about customers because the tools they use to manage digital assets and product data (traditional DAM and PIM systems) create unnecessary bottlenecks impeding the flow of personalized experiences to their target audiences. At Impira, we not only quality-control products, but also dramatically reduce the number of manual steps required to release those products (and associated experiences) out into the wild, helping e-commerce teams act on their insights.
If you were at NRF but didn’t get a chance to say hello, drop us a note and we’d be happy to find some time to chat. Please reach out to us here or at email@example.com and connect with us on Facebook, Instagram, LinkedIn, and Twitter.