Artificial Intelligence

The Goldilocks problem: How dynamic pricing powered by AI can find the “just right” price

Many industries rely on dynamic pricing. On-demand transportation companies, such as Uber, match drivers and passengers more efficiently than a traditional flat taxi rate. While some customers lament surge pricing, raising prices at peak periods of demand boosts supply because more drivers are incentivised to pick up passengers. Some motorways rely on dynamic pricing to set tolls, which rise and fall with demand to speed traffic flow. Online retailers, such as Amazon, continually adjust prices based on competitors, time of day, conversion rates, and more. Hotels leverage dynamic pricing to adjust the price of rooms based on the supply of vacant rooms, expected demand, and how far out a reservation is booked.

Despite the wide-ranging precedent, at many companies there is often an inertia to experiment with new pricing approaches. Sometimes this is because business leaders are unsure of what else they can do, or they lack granular visibility of pricing performance data. Other times there isn’t a mechanism to model and test microdecisions. These blockers make it impossible to solve the Goldilocks problem to find the price that is just right.

For many companies, dynamic pricing can close that gap. Dynamic pricing relies on frequent price adjustments to match supply and demand. Prices can vary daily or even every few minutes. The most effective approach is to continuously learn customer behaviour when setting prices. This is best achieved with machine learning.

At Elula, machine learning and mathematical optimisation power our dynamic pricing engine. Machine learning techniques are used to segment customers, develop price sensitivity curves for each segment, and predict demand and likelihood that inventory will be sold. Pricing optimisation algorithms then evaluate billions of options and in real-time recommend the optimal price to offer a specific customer at a specific time to align with business priorities (see diagram 1).

 

Case study: Dynamic pricing in logistics

Like a hotel that can’t sell its unsold rooms the next day, logistics companies that have dedicated space on aircraft hold perishable inventory. In a constrained supply environment, dynamic pricing can improve margins and unit costs for optimal profit (see diagram 2).

 

Since different customers are willing to pay different prices for the same product, price sensitivity curves are developed for each customer segment and include how factors influence the likelihood to buy, such as seasonality, day, industry, and urgency trends (see diagram 3). Dynamic pricing is driven by price elasticity curves that continuously optimise profit or any other business outcome, such as market share.

 

Dynamic pricing can also increase volume on days of lower demand through nudging customers that are price sensitive. This decreases unit costs and increases profit (see diagram 4).

 

Dynamic pricing drives many business benefits

Dynamic pricing maximises revenue, profit, conversion, market share or any other desired business outcome. Blending predictive and prescriptive machine learning techniques empowers leaders to act with precision and confidence, and in real-time. Instant pricing recommendations cut the long lead time from manual, bespoke pricing decisions and enable employees to instead focus on customer-facing high value activities. In addition, an easy-to-use interface, such as Elula’s, can provide transparency to why the optimisation algorithm recommended a specific price.

Evaluate billions of options and in real-time recommend the optimal price to offer a specific customer at a specific time

More Case Studies

  • Data Visualisation

    Stop flying blind: how world-class data visualisation can drive unprecedented business…

  • Advanced Analytics

    Transform how your business operates and empower leaders to respond to any…

  • Artificial Intelligence

    The Goldilocks problem: How dynamic pricing powered by AI can find the “just right”…