If you've ever seen a steep increase in the fare for an Uber to the airport on a Friday, or you've checked an item's cost on Amazon, only to see it has changed hours later, you might have experienced algorithmic pricing.
That's the practice of using algorithms to automatically adjust the price of goods or services based on factors such as demand, competitor pricing, inventory levels, or data about the customer.
While such pricing practices can squeeze out extra profit, they can also carry a marketing risk if not carefully implemented, according to Gizem Yalcin Williams , assistant professor of marketing at Texas McCombs. In 2012, Uber was widely criticized for raising ride prices during Hurricane Sandy . More recently, customers have expressed outrage over concert ticket surge pricing .
In a new paper, co-written with an interdisciplinary group of 12 other researchers, Williams examines algorithmic pricing and the challenges companies can face when integrating it with their other objectives. The researchers offer some preliminary dos and don'ts for aligning pricing with marketing strategy, regulations, and avoiding customer backlash.
One potential factor in customer backlash, Williams says, is driven by feelings of unfairness.
"Let's say that I just got myself something from Amazon, for my dorm, and then a couple of days later, I saw that the price changed," she says. "I now feel like I overpaid for it, regardless of how good the product is."
By the same token, seeing a price increase later might trigger elation, she says. "If I feel like I bought it at a lower price, I feel like I was smart."
When Prices Get Personal
If pricing sometimes feels a bit more personal when algorithms are involved, Williams says, that's because it is.
In addition to taking supply or production costs into account, companies increasingly use customer-level data to make pricing decisions, often with the help of artificial intelligence.
The exact data that go into the algorithm might not be always known, Williams says. "But what if the price I receive is different than others because of my own data, such as my shopping history, demographics, or location? Shoppers might react to the same price differently, depending on which data they think affected the price set by the company's algorithm."
Besides eroding customer loyalty, companies can face regulatory or legal attention when dynamic or surge pricing goes awry. Last year, the grocery chain Kroger was scrutinized by members of Congress over its plans to introduce algorithmic pricing at its stores.
As part of its research, Williams' team surveyed pricing managers and conducted in-depth interviews with five strategic-pricing experts. They offered several pieces of advice.
- Companies should be aware of how accepting their customers are — or are not — of dynamic pricing to avoid potential reputational damages.
- Opening the "black box" and increasing transparency about how algorithms work can help managers and employees adopt and oversee them effectively.
- Companies need guardrails to make sure they can effectively and carefully navigate the competitive and regulatory environment.
Williams sees the paper as raising timely and important questions for future research on consumer attitudes about algorithmic pricing. One takeaway, she notes, is clear: Many companies slap the AI label on their operations, to cut costs or boost efficiency, without comprehensive planning for its design, integration, and monitoring.
"Managers need to be deliberate about when, where, and whether to integrate AI into their operations," she says. "And even when decisions are automated, it's critical to have mechanisms that keep humans in the loop."
Algorithmic Pricing: Implications for Marketing Strategy and Regulation " is published in International Journal of Research in Marketing.
