Surge pricing is a variant of dynamic pricing (also known as variable pricing). The dynamics of prices means that prices can now change much more frequently and vary across customers, time and place at ever higher resolution; a price surge or hike at peak moments in demand can be described as an outcome of dynamic pricing. Surge pricing received great attention due to Uber’s application of this strategy, and not least because of the controversial way that Uber implemented it. But dynamic pricing, and surge pricing within it, is a growing field with various forms of applications in different domains.

A price surge is generally attributed to a surge in demand. In the case of Uber, when the number of customer requests for rides (‘hailing’) critically exceeds the number of drivers available in a given geographic area, Uber enforces a ‘surge multiplier’ of the normal (relatively low) price or tariff (e.g., two times the normal price). The multiplier remains in effect for a period of time until demand can be reasonably met. The advantages, as explained by Uber, are that through this price treatment (1) drivers can be encouraged to join the pool of active drivers (i.e., ready to receive requests on Uber app), as well as  pulling drivers from adjacent areas; and (2) priority can be given to a smaller group of those customers who are in greater need of prompt service and are willing to pay the higher price. Consequently, waiting times for customers willing to pay the price premium will be shorter.  (Note: Lyft is applying a similar approach.)

There are some noteworthy aspects to the modern surge pricing. A basic tenet of economic theory says that when demand surpasses the supply of a good or service, its price will rise until a match is reached between the levels of demand and supply so as to ‘clear the market’. Yet the neo-classic economic theory also assumes that the equilibrium price applies to all consumers (and suppliers) in the market for a length of time that the stable equilibrium prevails; it does not account well for temporary ‘shocks’. Proponents of surge pricing argue that this pricing strategy is an appropriate correction to a market failure caused by short-term ‘shocks’ due to unusual events in particular places. There is room in economic theory for more complex situations that allow for price differentials such as seasonality effects or gaps between geographic regions (e.g., urban versus rural, central versus peripheral). Still, seasonal prices are the same “across-the-board” for all; and regions of different geographic markets are usually well separated. On the other hand, in surge pricing, and in dynamic pricing more broadly, it is possible through advanced technology to isolate and fit a price to a very specific group of consumers in a given time and space.

One of the concerns with surge pricing in ride e-hailing is that the method could take advantage of consumers-riders when they have little choice, cannot afford to wait too long (e.g., hurry to get to a meeting or to the airport) or cannot afford a price several times higher than normal (e.g., multipliers of more than 5x). The problem becomes more acute as surge pricing seems to ‘kick in’ at worst times for riders, when they are in distress [1a](e.g., in heavy rain, late at night after a party). The method seems to screen potential riders not based on how badly they need the service but on how much they are willing to pay. The method may fix a problem for the service platform provider more than for its customers. Suppose hundreds of people are coming out at the same time from a hall after a live music concert. If the surge multiplier shown in the app at the time the prospect rider wants to be driven home is too high because of the emerging peak in demand, he or she is advised to wait somewhat longer until it slides down again. How long should riders wait for the multiplier to come down? Often enough, so it is reported, it takes just a few minutes (e.g., minor traffic fluctuations). But in more stubborn situations the rider may be able to catch a standard taxi by the time the multiplier declines, or if the weather permits, walk some distance where one can hail a taxi or get onto another mode of public transport.

Another pitfall is reduced predictability of the occurrence of surge pricing. Consumers know when seasons start and end and can learn when to expect lower and higher prices  accordingly (though it used to be easier thirty years ago). In public transport, peak hours (e.g., morning, afternoon) are usually declared in advance, wherein  travel tariffs could be elevated during those periods. Since surge pricing is based on real-time information available to the service platform provider, it is harder to predict the occasions when surge pricing will be activated, and furthermore the extent of price increase. Relatedly, drastic price changes (e.g., due to high frequency of updates, strong fluctuations) tends to increase the uncertainty for service users [1b].

The extent of price surge or hike is a particular source of confusion. Users are notified before hailing an Uber driver if surge is on, and a surge multiplier in effect at that time should appear on the screen. The multiplier keeps being updated on the platform. It is sensible, however, for the multiplier to stay fixed for an individual rider after the service is ordered. Thus the rider can make a decision based on a known price level for the duration of the ride (or an estimate of the cost to expect). Otherwise, the rider may be exposed to a rising price rate while being driven to destination — but the rider should also benefit if the multiplier starts to slide down (or entering another area where surge is off). The first scenario resembles more a situation of bidding whereas the latter scenario looks more like gambling. Stories and complaints from Uber users reveal recurring surprises and unclarity about the cost of rides (e.g., claims the multiplier was 9x, a ride of 20 minutes that cost several hundreds of dollars, a claim the multiplier dropped but the total price did not go down in accordance). Users may not pay attention sufficiently to the multiplier before hailing a ride, do not comprehend how the pricing method works, or they simply lose track of the cost of the ride (i.e., the charge is automatic and appears later on the user’s account).

Discontent of customers may also be raised by a sharp contrast experienced between the relatively low normal price rate (e.g., compared with a standard taxi) and the high prices produced by surge multipliers [1c].  A counter argument contends that the price hikes or surges allow for low rates at normal times by subsidising them [2]. More confusion about Uber’s pricing algorithm could stem from reports on additional factors that the company might use as input (e.g., people are more receptive of surge prices when the battery of their mobile phone is low, and customers are more willing to accept a rounded multiplier number than a close non-rounded figure just below or above it (MarketWatch.com, 28 December 2017).

• Not even a strategy of surge pricing appears to be completely immune to attempts of manipulation. It was revealed in 2019 that drivers with Uber (and also Lyft’s) have tried to game the surge mechanism. The ‘trick’ is to turn off the app at a given time in a coordinated manner among drivers, let the surge multiplier rise, and then turn on the app again to gain quickly enough from the higher rate as long as it prevails. The method seems to have been used especially at airports in anticipation of incoming passengers, based on the knowledge of drivers of several flights scheduled to land during a short interval. The motivation for taking this action: the drivers claim they are not paid enough at normal times by the platform operators (BusinessInsider, 14 June 2019).

Uptal Dholakia, a professor of marketing at Rice University (also see [1]), suggested four remedies to the kinds of problems described above. First, he advised to set a cap (maximum) on surge multipliers and notify customers more clearly about them (greater transparency). In addition, he recommended curbing the volatility of price fluctuations and communicating better the benefits of the method (e.g., reduced waiting times). Dholakia also raised an issue about a negative connotation of the term ‘surge’ that perhaps should be replaced in customer communications [3].

Various forms of dynamic pricing, including surge pricing, are already utilised in multiple domains. It is noted, for instance, that the strategy of Uber was not initiated to resolve problems of traffic congestion; ‘surge’ may be activated as its result but the purpose is to resolve the interruptions that congestion may cause to the service. For dealing with traffic congestion and overload in roads, other types of surge pricing are being used by public authorities. First, a fast lane is dedicated on a highway or autoroute (e.g., entering a large city) for a fee — the amount of ‘surge’ fee is determined by the density of traffic on the other regular lanes. Drivers who wish to arrive faster should pay this fee that is displayed on a signboard as one approaches entry to the lane (a few moments are allowed to decide whether to stay or abort). Second, a congestion fee, which could actually be a variable surge fee, may be imposed on non-residents who seek to enter the municipal area of a city at certain hours of the day.

As indicated earlier, public transportation systems in large cities may charge a higher tariff during peak or rush hours. The time periods that a raised tariff applies are usually declared in advance (i.e., they are fixed). Peak and off-peak rates may apply to different types of travel fares. The scheme is employed to encourage passengers who do not really need to travel at those hours to change their schedule and not further load the mass transportation system. There is of course a downside to this approach for passengers who must travel on those hours, such as for getting to work (employers who cover travel expenses should set the amount according to the cost of the more expensive rate). Using surge pricing in this case would mean that passengers cannot tell for certain and in advance when a higher tariff applies, but the scale of ‘surge prices’ can be pre-set with a limited number of ‘steps’, and thus reduce resentment and opposition.

Other types of dynamic (variable) pricing involve strong technological and data capabilities, including information on demand at an aggregate level and customer preferences and behaviour (search, purchase) at the individual level. A company like Amazon.com keeps updating its prices around the clock based on data of demand for products sold on its e-commerce platform. A more specific type of dynamic pricing entails the customisation of prices quoted to individual users-customers (i.e., different prices for the same book title offered to different customers). The approach maintains that a higher price could be set, for instance, for books in a category in which the customer purchases books more frequently and even based on search for titles in categories of interest. This form of price customisation is debatable because it aims to absorb a greater portion of the consumer’s value surplus (i.e., how much value a consumer assigns to a product above its monetary price requested by the seller), raising concerns of unfairness and discrimination. The risk to sellers is of making products less worthwhile to consumers to buy at the higher customised prices. (Note: Amazon was publicly blamed of using some form of price customisation in the early 2000s after customers discovered they had paid different prices from their friends; however the practice has not been banned and it is suspected to be in use by companies in different domains.)

• Take for example the air travel sector: Airlines may use any of these methods of variable pricing: (a) Offering the same seat on the aircraft at different price levels (‘sub-classes’) depending on the timing of reservation before the scheduled flight: the earlier a reservation is made, the lower the price; (b) Changing fares for flights to different destinations based on fluctuations in demand for each destination and time of flight; (c) There are claims that airlines also adjust upwards the fares on flights to destinations that prospect travellers check more frequently in the online reservation system.

More companies in additional sectors are expected to join by applying varied forms of dynamic pricing. Retailers with physical stores are expected foremost to use dynamic pricing more extensively to tackle the growing challenges they face particularly from Amazon.com in the Western world (e.g., supermarkets will employ digital price displays that will allow them to change prices more continuously during the day and week according to visitor traffic levels). Restaurants may set higher prices during more busy hours at their premises, and hotels are likely to vary their room rates more intensively, taking into consideration not only seasonal fluctuations but also special events like conferences, festivals and fairs (e.g., see “The Death of Prices”, Axios, 30 April 2019).

Dynamic pricing, and surge pricing in particular, is the new reality in pricing policy, with applications getting increasingly pervasive. As technological and analytical capabilities only improve, the pricing models and techniques are likely to be enhanced and become furthermore sophisticated. Moreover, methods of artificial intelligence will improve in learning patterns of market and consumer behaviour, expected to enable companies to set prices with greater specificity and accuracy. At the same time, businesses need to take greater caution not to deter their customers by causing excessive confusion and aggravation. The question then becomes: What bases of discrimination — among consumers, at different times, and in different locations — would be considered fair and legitimate? This promises to be a major challenge for both enterprises that set prices and for the consumers who have to judge and respond to the dynamic prices.

Ron Ventura, Ph.D. (Marketing)

Notes:

[1a-c] “Uber’s Surge Pricing: Why Everyone Hates It?”, Uptal M. Dholakia, Government Technology (magazine’s online portal), 27 January 2016

[2] “Frustrated by Surge Pricing? Here’s How It Benefits You in the Long Run”, Knowledge @Wharton (Management), 5 January 2016. A talk with Ruben Lobel and Kaitlin Daniels at Wharton Management School at the University of Pennsylvania.

[3] “Everyone Hates Uber’s Surge Pricing — Here’s How to Fix It”, Uptal M. Dholakia, Harvard Business Review (Online), 21 December 2015