Peter Andrews is a managing director in Alvarez & Marsal’s Corporate Performance Improvement team.
With national grocery chains shifting to electronic price tags, the ability for retailers to deploy AI-enabled dynamic pricing may soon be commonplace as companies look to be more nimble amid rapidly shifting market forces. Improved responsiveness will not only help companies stay steady but also grow their bottom lines. But moving too quickly may come at the expense of their most loyal customers and attract regulatory attention.
Rather than investing effort in periodic price models that become obsolete soon after implementation due to tariffs, supply chain shifts and swings in material costs, machine learning algorithms enable companies to use real-time data to inform prices. AI can ingest vast amounts of data to determine pricing recommendations, including inputs like material costs, competitor pricing, customer demand, predicted demand, inventory levels and potential delays or disruptions in the supply chain. Without the lag of constantly updating static pricing, AI-enabled pricing models allow companies to better capture margin on their products by automatically rolling out real-time prices.
Data-informed pricing isn’t a new concept: The aviation and hospitality industries have long used these models to appropriately price hotel rooms and plane seats. Dynamic pricing algorithms have helped airlines reduce empty seats, increasing their load factor from 72% in the early 2000s to over 80%. Consumers have accepted that ticket prices may shift depending on when they’re purchased, largely because dynamic pricing has been mutually beneficial: Ticket prices are, on average, lower, while airlines have increased revenue per available seat mile. And they continue to innovate. This year, a large domestic airline announced it was now leveraging AI in its pricing algorithms, resulting in higher profitability; however, it has been facing some scrutiny from lawmakers.
As AI-enabled pricing enters new industries, the potential harms of AI-enabled dynamic pricing are becoming more tangible for consumers and regulators alike.
For all the benefits that AI-enabled dynamic pricing offers to businesses, the technology also creates significant concerns about algorithmic bias and price gouging. For example, personalized automated pricing could theoretically give consumers different prices for the same exact product at the same exact time, based on an algorithm that calculates just how much a given individual is willing to pay. Or, without proper guardrails, automated pricing could cause the cost of water or plywood to soar in the face of a hurricane as demand spikes.
The key for industries experimenting with AI-based dynamic pricing is to understand their customers’ expectations. A manufacturing company in the B2B space, for example, could have an easier time selling customers on the idea that dynamic pricing flexes both ways — saving them money when costs are low and protecting the manufacturer’s earningswhen input costs jump up. A company with a more generalized client base will have a much harder time making that sell, as the fast food industry learned recently as it toyed with the idea of “surge pricing.”
Charging full-bore into fully AI-powered pricing models is a risk, not just for a company's relationship to its customers, but also because of regulatory scrutiny. The FTC, for example, has started to watch for companies employing "surveillance pricing," the practice of utilizing customer data to dynamically set prices. Last year, New York state put a law into effect that requires businesses to disclose when prices were derived from an algorithm that utilized personal data. This law, along with bills introduced in other states, reinforces the need for increased vigilance when it comes to pricing algorithms.
Recently, an online delivery service faced scrutiny for using AI-enabled pricing to charge different prices on the same items to different customers within the same store. The revelation sparked media backlash, and the company announced that they immediately halted the practice. Contrast that to another example where several large grocery chains in Europe adjust prices dynamically throughout the day. The difference here is that the prices predominantly move down and do not target specific customers, creating a disadvantage. Rather, they use the price changes to be competitive and reduce inventory spoilage.
Deliberate, well-tested adoption is key to safeguarding customer trust and minimizing regulatory scrutiny. AI-powered dynamic pricing doesn’t mean that pricing becomes a set-it-and-forget-it growth tactic, but rather a thoughtful, brand-aligned strategy to optimize revenue and limit cost exposure.
As with all AI adoption in businesses, it is important that humans remain in the driver’s seat, monitoring prices to ensure customers don’t suffer unintended consequences or algorithmic bias. AI pricing is not inherently a risk to consumers; the introduction of digital price tags in grocery stores, which could enable more rapid dynamic pricing, has not been associated with any significant price hikes. Prices changed by just 0.0006% after the introduction. But unmonitored AI pricing is a risk — one that can be an open invitation for regulators to intervene, especially if its use creates the appearance of collusion, feeding price increases through more price increases as dueling AI systems keep pace with one another to drive higher profits.
Dynamic pricing is an essential tool that helps businesses better respond to a market moving faster than ever, but retaining customers in the transition depends on a well-thought-out strategy. AI can identify when and where to adjust pricing faster, but human judgment — and the right guardrails — remain essential.
