What is dynamic pricing in 2019?
Airlines are often acknowledged as the trailblazers of dynamic pricing, since they started to offer the same product at multiple different price points in the 80s and 90s. However, over the last decade other online retailers have taken dynamic pricing to a whole new level, leveraging new technologies and utilizing new data-driven strategies.
While dynamic pricing is often thought of as the ability to offer a product at different price points, it is really a set of evolving pricing strategies that enables the price to be dynamically determined in real-time based on various criteria.
While many airlines have a relatively high level of maturity in how they price their tickets, their level of pricing maturity for some of their new product offerings is much less developed. This is mainly due to a lack of capabilities in their current systems. These were designed for traditional airline ticket sales and don’t typically lend themselves to the pricing of multiple digital products and service offerings.
Even for airline ticket pricing, airlines are beginning to explore ways to leverage new technologies to enhance current demand-based pricing and experiment with personalized pricing techniques.
The problem with traditional dynamic pricing
Traditionally dynamic pricing was calculated based on a retrospective analysis of sales data across various sales channels. This enabled the dynamic pricing system to forecast market demand by observing cyclical or seasonal patterns. This technique was effective in comparison to static pricing but had some shortcomings.
Estimates of market demand lagged the actual demand in the market. Sales data, while a good estimator of market demand, does not capture all the factors that might influence demand at any given time.
It also relied heavily on cyclical and seasonal patterns which are reliable when the market is stable. But these become less reliable when the market and/or competition become more volatile and hence less predictable.
Modern dynamic pricing solutions use the latest technologies and algorithms to generate real-time dynamic prices that leverage multiple sources of data to more accurately estimate market demand.
Performing these calculations in real-time enables the current market demand to be used to determine the current offered price. The elimination of the time lag between changes in market demand and the reciprocal pricing changes, combined with the more accurate estimation of current market demand, provides a competitive advantage to those who are first to embrace the latest innovations.
Technological advancements changing dynamic pricing
With technology advancements in recent years, approaches to dynamic pricing have changed both in terms of what’s possible from a scalability point of view, and what machine learning algorithms can be leveraged to improve demand estimation.
For example, the emergence of distributed cluster computing platforms has enabled machine learning algorithms, which were previously prohibitively expensive, to be brought into mainstream applications.
These platforms allow the required computations to be broken down into smaller chunks that can then be processed in a massively parallel fashion, leveraging a modern cloud-based infrastructure. This means that proven academic techniques for machine learning such as supervised learning, unsupervised learning, reinforcement learning and deep learning, can now be leveraged in production systems in a timely and cost-effective manner.
To learn more, download the full white paper ‘Dynamic Pricing – Its Role in Digital Retail Thinking’ by Datalex Chief Innovation Officer Alan Dunne. This white paper continues to explain how machine learning can be applied to dynamic pricing, how leading retailers use dynamic pricing to drive demand, revenue and the customer experience, and what the future of dynamic pricing holds for airlines.