Re-inventing Revenue Management

February 28, 2014-

Martin Rivers explores the changes for back office processes that will come with the New Distribution Capability

The science of revenue management has completely changed during the last 25 years, mainly because of the advances in computer processor speed and the ability to apply more complex algorithms,” explains Surain Adyanthaya, Senior Vice President of Product Management of PROS, a sales analytics firm that develops software for airlines.

While these changes look set to continue, Adyanthaya insists that a more fundamental shift in revenue management is also under way. It is an evolution that when juxtaposed with the development of IATA’s New Distribution Capability (NDC), affords airlines greater presence in the marketplace through a more personalized booking process on indirect channels.

Segmentation has always been at the core of revenue management theory. But the ability of airlines to segment passenger types beyond relatively basic variables—the time until departure, for example, or the inclusion of a Saturday night in the itinerary—has been undermined by recent trends.

Unbundled revenue

The emergence of price-competitive low-cost carriers (LCCs) first disrupted the ability of revenue management forecasters to calculate the true value of their customers. As this new category of airlines stripped down service provision and replaced it with layer upon layer of optional add-ons—each attracting an ancillary surcharge—it became harder for airlines to predict passenger spending patterns. The true value of the customer was no longer self-evident at the booking stage.

“LCC pioneers really pushed the unbundling of the airline product,” Adyanthaya says. “Their argument was that it gives choice to the consumer and allows them to purchase only what they need, keeping the price low.

“But it also changed how we think about revenue management,” Adyanthaya continues. “Before the unbundling trend, we would essentially try to maximize the revenues based on the fares the passenger paid. Now we really have to think about the total overall revenue that the passenger will generate across their journey, including all these various ancillary products that they may purchase. It really changes the dynamics of how revenue management works.”

Two decisions

Dennis Corrigan, Vice President of Sales and Revenue Management at JetBlue Airways, agrees that traditional revenue management systems have been “relatively blind” to the added value that potential ancillary purchases bring to bookings.

He says that because of this limitation, airlines typically had to make “two independent decisions” about future revenue forecasts. The first is based on actual airfares and the second on predicted revenue from subsequent ancillary purchases. “Ticket revenue isn’t the only source of revenue we’re getting on the flight,” Corrigan stresses. “We need to be optimizing ancillaries too.”

But mathematical predictions about which groupings of passengers are likely to buy which ancillaries are not especially developed. Corrigan notes that airlines presently have little ability to correlate the two decisions. It is possible, for example, that higher airfare revenue from a late booking may be offset by a lower propensity to purchase ancillaries. Conversely, early-bird passengers who have benefited from the cheapest airfares may be more willing or likely to splash out on extras.

Nor is the time until departure the only variable at play. Passengers’ predisposition for ancillaries can be influenced by an array of factors, all of which should ideally be factored into the algorithms.

Adyanthaya notes how some airlines currently charge the same flat fee for extra legroom on a Saturday morning flight as on a Monday morning flight. But does it have the same value for the customer? A business traveler might well pay more on a Monday morning. And it may not pay to price leisure travelers out of a flight on a route where duty free sales are strong.

“There’s not much science-based price differentiation for many of these ancillary products,” he says. “Airlines should charge the right price at the right time to the right customer. We believe that as science and technology evolve, that will be one of the big opportunities for the future.”

Big data solutions

The task of understanding the interplay of different variables and forecasting their impact on purchasing patterns rightly falls to algorithm coders. Choice modelling and behavioral analysis will play a key role in refining price optimization in the era of unbundled tickets.
But the industry is collectively responsible for creating a landscape that fosters the development of this science. Only by improving access to and utilization of passenger data can airlines unlock the true potential of dynamic revenue management theory.

Progress is constantly being made. Real-time optimization platforms incorporate an ever-growing array of data, conducting immediate analyses of availability requests before offering prices.

Yet there remains an undeniable gap between the functionality of direct channels (airline websites) and indirect channels (global distribution systems–GDSs and online travel agents–OTAs). “There’s much less capability to custom-create the offer for the prospective traveler on indirect channels, just because of the nature of how all the data is shared,” Adyanthaya complains.

Whereas direct channels can be readily improved through bespoke Customer Relationship Management (CRM) systems, an industry-wide standard is required before the legacy IT systems used on indirect channels can be replaced. Pilot schemes for IATA’s NDC platform, which will use up-to-date XML technology, are underway with American Airlines, Air New Zealand, and Swiss International Air Lines.

“Indirect channels have a somewhat outdated, rough kind of segmentation, because the industry started distribution in a pre-internet age,” says Karl Isler, Head of Operations Research and Strategy at the Revenue Management and Pricing department of Swiss International Air Lines.

“NDC will be much more differentiated so you will have more gradual variation. The same type of segmentations would apply but on a much more refined level. NDC gives us very good opportunities to do what we do today but do it much better; to understand exactly what the customer is searching for.”

At the heart of NDC’s proposition lies the personalization of the booking process. The platform will allow OTAs and GDSs to present the full array of ticket offerings in the marketplace, breaking down searches by ancillary services and providing in-depth detail about each airline’s product.

To the consumer, this will make comparisons between travel products more transparent. Travelers using an OTA— which today generate search results based solely on price and date—will have the full spectrum of product details at their fingertips. The qualitative differences between low-cost, hybrid, and full-service carriers will be self-evident, as will the minutiae of product variations (everything from seat sizes to menus to in-flight entertainment systems).

But more advanced functionality will also benefit airlines. By inviting consumers to voluntarily share data on their travel history and personal preferences, tailored products can be proactively marketed to the individual searching for a flight.

“NDC will generate a much more intelligent, personalized offer,” Adyanthaya explains. “Airlines have always spoken about the lifetime value of a customer and the need to really understand the customer at a higher level. NDC allows the airline to know what it should offer the customer and how it should treat the customer. It’s about really segmenting customers and then optimizing those market segments.”

Intelligent re-bundling

The precise form of such personalization remains open for debate. In theory, Corrigan says a 150-seat aircraft could have 150 price points. In practice, however, algorithms will still need to bunch together revenue streams.

“It’s not possible to do forecasting on the individual level,” Isler notes, also emphasizing the need to aggregate predicted revenue. “I would see personalization rather in the customer relation dimension: frequent flyer tier members and so on. The abstract revenue management part has to keep to certain tangible elements.”

Even as tailored products create the aura of personalization then segmentation will lend a degree of certainty to revenue management. To this end, intelligent re-bundling has already emerged as the favored strategy. NDC will expand this to indirect channels, using voluntary customer information to create tickets that accommodate a given traveler’s preferences. Entertainment bundles might be pre-selected for family vacation bookings, for example, while business travelers might be offered airfares with WiFi and extra legroom.

Such products will be constantly refined as they are tested in the marketplace. Algorithms will be optimized over time, with sales agents marketing novel combinations of ancillaries through NDC. As well as promoting their services more effectively, airlines should gradually regain more control of revenue forecasts.

“At the heart of it, the number of packages, products or bundles I’m selling is still going to be limited to the number of seats on the aircraft,” Corrigan concludes. “It’s still about the Expected Marginal Seat Revenue of keeping a seat open versus closing it. But demand forecasts will have to be based on more than just ticket price.”


Houston Chronicle: People in Business – February 20, 2014


Chief Marketer: Moneyball and Marketing: Challenging Conventional Data Wisdom