Dynamic pricing, a practice started by American Airlines in the 1980s, has now become a common marketing discipline for many corporations across industry sectors. From airlines, hotels, and entertainment events to perhaps the most well-known e-retailer, Amazon, these companies have been using dynamic pricing to improve profitability relative to rapid changes in supply and demand.
Personalization: Using AI to deliver optimal customer experiences, in real-timeAs technologies for machine learning and artificial intelligence become more advanced and the dimensions of available data expands, dynamic pricing is going beyond its traditional inventory management function enabling companies to deliver optimal customer experiences, in real-time. In essence, pricing is becoming dependent on the ability to make offers which continuously adjust to changing consumer behavior and preferences while also responding to organizational inventory and profit requirements, as well as other external pricing influences. Today, enterprises are able to marry rich data sets with sophisticated pricing models and apply advanced analytics and machine learning techniques to produce pricing alternatives across thousands of products SKU’s. These offers are also uniquely tailored to the individual consumer dynamically, at the point of engagement.
Machine models enable humans to apply business judgement across vast amounts of dataThe application of technology alone, however is insufficient. All pricing actions need to embody the brand value and ultimate profitability of the business, this requires human guidance and oversight in the form of constraints and outcome objectives. Balance is achieved when the machine models become enablers for the human to apply business judgement across an ever expanding and changing source of data.
Going back to the airline industry as an example; airlines operate in an extremely competitive environment with high fixed costs and razor thin margins. To complicate matters further, airlines are subject to a wide array of unpredictable events that frequently affect operations, including demand, inclement weather, crew illness and aircraft maintenance to name a few. Combine these with consumer loyalty, route availability, schedule changes and seasonal demand, dynamic pricing begins to take on a rather complex decision process. Ongoing insight from available data sources can be the key to unlocking a dynamic pricing solution.
Gartner iterates that the, “Improved customer experience from real time, contextualized interactions supported by AI and machine learning is a top trend for CEOs… Digital commerce will also leverage AI and machine learning to predict customers' needs, and proactively offer relevant products and services.” Directing that capability beyond just consumer behavior, and including external factors that are foundational to strong business returns is now within reach across multiple industries beyond e-commerce. The computational power of artificial intelligence in the hand of the human guided pricing decision is no longer the future, but the new norm.
By example, the auto finance industry is using advanced analytics to predict customers' needs and deliver highly personalized financing options while taking into consideration a wide array of variables. Such variables can include incentives from manufacturers, the value of the trade-in vehicles, and individual borrowers’ preferences, such as the down payment they can afford or the payment terms they prefer. By analyzing multiple data sets, lenders can offer customers personalized options to quickly win their business.
The ability to process, monitor and analyze information in real time so that consumers are provided with contextualized and highly personalized pricing based offers is a commerce imperative. A new frontier of dynamic pricing is good for the consumer and good for the bottom line. I personally love how AI and ML are providing a win-win in this use case.
Originally published on Forbes.com.
For more insight into price optimization, check out this recent FICO Blog post: