Summary by James R. Martin, Ph.D., CMA
Professor
Emeritus, University of South Florida
Pricing Decisions Main Page |
Surveys Main Page
Pricing algorithms are designed to help companies determine optimal prices on a near real-time basis. They are computer programs that combine artificial intelligence and machine learning to analyze variables such as supply and demand, competitor pricing, and delivery time. Companies in many industries such as advertising, e-commerce, entertainment, insurance, sports, travel, and utilities have used this approach to pricing. However, although dynamic pricing algorithms can enhance revenue, they can also trigger unfavorable perceptions about the company and the company's products. The purpose of this article is to provide real world examples of algorithmic pricing and ways it created a benefit or harmed the associated brand, and to examine how proper oversight and management of pricing algorithms can limit the potential for misuse, and instead enhance both the company's revenue and image.
Examples of Algorithmic Pricing that Produced Unfavorable Results
Some examples of how algorithmic pricing produced unfavorable results include how Uber's algorithm increased rates by more than 200% during a 2017 terrorist attack in London, and produced similar price increases in a 2016 bombing in New York City, a 2017 taxi drivers strike, and a 2020 mass shooting in Seattle. Another classic example is Coca-Cola's temperature-sensitive vending machines that increased the price of a beverage on a hot day. Both companies received a considerable amount of criticism for this type of algorithmic pricing.
The Psychological Impact of Algorithmic Pricing
Root Insurance sells automobile policies using a dynamic pricing approach based on a driver's driving record. Drivers are offered a smartphone app that measures their driving behavior and calculates their insurance premiums based their individual safety scores. This is an effective example of how dynamic pricing can improve customer relationships. Customers understand why they are offered a specific price.
Four Recommendations
The authors provide four recommendations for obtaining favorable results from pricing algorithms:
1. Determine an appropriate use case and narrative.
2. Designate a pricing algorithm owner.
3. Set and monitor pricing guardrails.
4. Override the algorithms when necessary.
Determine an Appropriate Use Case and Narrative
The Swedish furniture retailer IKEA developed a dynamic pricing algorithm to reward customers for traveling to their store. The algorithm uses Google Maps Timeline readouts to factor in the distance traveled, along with other variables to determine the price customers pay. This provides an incentive to travel long distances to purchase items from IKEA. Their algorithm rewards customers rather than penalizing them for shifts in supply and demand as in some of the cases mentioned above.
Designate a Pricing Algorithm Owner
In 2019 United Airlines eliminated the mileage tables for frequent fliers and replaced them with an algorithmic pricing model. They explained why it was necessary to base award travel on supply and demand and how customers could benefit from the change. They also delegated the management of the algorithm to the group that supervised the loyalty program to help monitor and respond to challenges with customer relationships. This example emphasizes that companies should not allow the algorithm to control the math and the messaging. Algorithms that emphasize supply and demand lack empathy and a long-term perspective related to customer relationships and loyalty. Empowering a team to monitor the effects of the algorithm can limit the potential for unfavorable results.
Set and Monitor Pricing Guardrails
To limit the typical poor experience at its theme parks (long lines for rides, food, and restrooms) Walt Disney World in Orlando initiated algorithmic pricing in 2018. It increased multiday-ticket prices, but decreased the price of tickets for off-peak dates and encouraged customers to plan trips during off-peak periods. This helped produce a more even flow of customers, smoothed staff and resource demands, and improved the typical customer experience. They also set guardrails for single-day ticket prices across their theme parks so that they would not vary significantly based on the time of year a customer chose to visit, regardless of demand.
The authors recommend three areas for closer collaboration across functions to obtain insights from pricing algorithms; 1. Controlled pricing experimentation to understand how customers react to offerings and price changes. 2. Monitoring price changes to measure the effects on customer loyalty or brand reputation. 3. Determining whether the company's product development, branding, positioning, and pricing work together to promote the firm's strategic objectives.
Override the Algorithms When Necessary
Pricing algorithms can analyze customer data and other information to generate optimal prices for any customer, but there is a conflict between earning money and earning customer goodwill. This produces a challenge for companies that should be monitored by a clear owner and managed, tweaked, or temporarily suspended when necessary. All pricing algorithms are communicating to customers, and companies need to learn how to control that message. This is accomplished by developing a proper use case and narrative for implementing dynamic pricing, assigning an owner to monitor pricing guardrails, and empowering the owner to override the automation when needed. Following these recommendations will allow for optimized pricing without damaging customer loyalty and the company's reputation.
________________________________________
Related summaries:
Abel, R. 1978. The role of costs and cost accounting in price determination. Management Accounting (April): 29-32. (Summary).
Govindarajan, V. and R. N. Anthony. 1983. How firms use cost data in price decisions. Management Accounting (July): 30-31, 34-36. (Summary).
Hinterhuber, A. 2008. Customer value-based pricing strategies: Why companies resist. Journal of Business Strategy 29(4): 41-50. (Summary).
Hinterhuber, A. and S. Liozu. 2012. Is it time to rethink your pricing strategy? MIT Sloan Management Review (Summer): 69-77. (Summary).
Hughes, S. B. and K. A. Paulson Gjerde. 2003. Do different cost systems make a difference? Management Accounting Quarterly (Fall): 22-30. (Summary).
Martin, J. R. 2022. A note on pricing strategies. (Note).
Porter, M. E. 1980. Competitive Strategy: Techniques for Analyzing Industries and Competitors. The Free Press. (Summary).
Shim, E. and E. F. Sudit. 1995. How manufacturers price products. Management Accounting (February): 37-39. (Note).
Shim, E. D. and R. Lim. 2022. A survey of U.S. firms' pricing strategies and costing methods. Cost Management (May/June): 15-19. (Note).
Thurston, K. L. D. M. Keleman and J. B. MacAarthur. 2000. Providing strategic activity cost information: Cost for pricing at Blue Cross and Blue Shield of Florida. Management Accounting Quarterly (Spring): 4-13. (Summary).