Eye on AI - November 15th, 2021
Welcome to Aigora's "Eye on AI" series, where we round up exciting news at the intersection of consumer science and artificial intelligence!
This week, we’ll be looking at how AI will soon touch upon every aspect of retail, including logistics, fraud, personalization, and beyond.
The Ways In Which AI Will Soon Touch Every Part of Retail
It’s no secret that AI has been slowly permeating into every part of our daily lives. Less obvious is how that transition is affecting retail. As Forbes contributor Richard Kestenbaum notes in his latest article, “Like It Or Not, Artificial Intelligence Is Coming To Every Part Of Retail”, that transitioning is happening. And it’s happening much faster than you think.
To begin his article, Kestenbaun points to a relatively simple AI application in retail now being implemented in refrigeration management. According to the EPA, a supermarket of 50,000 square feet will spend roughly $150,000 per year on gas and electricity for refrigeration. Most refrigeration systems are currently managed by people. By using AI to more efficiently manage refrigeration, supermarkets could save upwards of $60,000 in a single year, over 40% of the total cost –– that’s the savings of a single store. Publix is currently rolling out a new AI system called COI Energy Services to manage the refrigeration at all locations. Add them up, and AI will potentially save them millions every year.
That’s just one application. The beauty of AI is that it can be applied to every retail division and in just about as many ways as you can imagine, even tasks that require more of human touch such as branding. Unlike refrigeration, which looks specifically at gas and electrical usage, branding is unique in that to do it effectively requires the concurrent analyzation of numerous data points from things like target audience, product delivery, packaging, promotion, distribution, customer experience (the list goes on and on) in real-time, then choose from a pool of limitless solutions.
“Until now, [brand] decision-makers relied on their instincts or ‘gut’ and while some people are gifted at brand management, most people aren’t,” writes Kestenbaum. “That’s why AI can help and competing against a brand managed with the help of artificial intelligence is more than anyone could handle. Everyone is going to have to do this.”
How can AI help with something as complicated as brand management, you might ask? The secret is something called ‘machine learning’ (ML), a version of artificial intelligence that uses algorithms to take in and analyze massive amounts of data in mere seconds.
“[Machine learning] ingests and analyzes that information and the analysis is the difference between artificial intelligence and everything else,” continues Kestenbaum. “The software learns over time and improves its own decision-making. It takes in far more information than any human being can and all day and night it analyzes its previous judgments to make better decisions and recommendations.”
In other words, information that would have previously taken brand managers years or lifetimes to analyze can now be delivered instantly, and in real-time, using AI. For instance, suppose inventory is high for product X, and a brand manager is tasked with finding ways to move excess inventory. But how, and how fast? Now suppose that the same brand manager is using AI to deliver her insights. By identifying that the warehouse is full and may not have room for the new shipment set to arrive next month, the AI might recommend freeing up inventory space with a promotion for product X both in-store and on the website. In addition, it could automatically send direct emails with the promotion, using on-brand content that already exists within an internal CMS (content management system), to those shoppers that have displayed a preference for product X through past shopping patterns. That’s just one example. The same kind of pattern identification can be used for things like finding the best product displays, choosing the most cost-effective or functional packaging, managing shipments, logistics –– the list goes on. Today, most retailers are only using AI for a limited number of applications, if at all.
“According to a study done by business monitoring company Anodot, retailers who are employing AI right now use it more for non-customer facing applications, like detecting fraud,” continues Kestenbaum. “But they expect to migrate the use of AI to more consumer-facing applications like personalization and visual search... If artificial intelligence is relevant to both brand management and the temperature of a freezer in a supermarket, then it will eventually be ubiquitous at every retailer.”
The problem in retail, as Kestenbaum astutely notes, is that no one wants to be first to adopt a trend and no one wants to be last. Brands look to other brands to lead the charge before hopping onto the bandwagon. But in terms of AI, with major retailers like Publix, Nike, BevMo, and others leading the charge, now is the time to follow suit.
That's it for now. If you'd like to receive email updates from Aigora, including weekly video recaps of our blog activity, click on the button below to join our email list. Thanks for stopping by!