The death of airline loyalty?

Last week, I wrote a post talking about how Digital in the airline industry could transform the boarding process.  Using off the shelf hardware and industry recognised business processes to dramatically redefine the customer experience in a disruptive and personalised way.

This week’s technology post goes into a lot of technical depth, but I think there are some fascinating implications about where airlines could take their use of technology, and also how consumers and travellers may adapt.

This all came about because last week, I read this post by Seth, the Wandering Aramean, on the latest development in the roll out of something the airline industry call the New Distribution Capability (NDC).

Before I dive into the technicalities, I’m going to explain how fares are sold today, and how the NDC aims to change things.

How airlines sell fares today

Today, there are two main ways to buy an air fare:

  • Direct via an airline’s own website
  • Via travel agents; either a corporate one, or an site like Expedia

Airlines generally prefer you to go direct and book via their websites as there are fewer middle-men to pay commission to and generally lower costs of sale for them.  Corporates still very much like to use an agent, as they can enforce policies, calculate rebates based on spend, and work out where their employees are at any one time.

The key ingredient in the travel agent route is something called a Global Distribution System (GDS).  This is essentially a connection from every airline to all the travel agents – for those that speak IT, it’s a very large middle-ware system.  There are four key GDS players in the market at the moment (Amadeus, Galileo, Sabre and Worldspan) and they’re all interconnected.

They act as a clearing house or Amazon-like service where travel agents can buy airfares from nearly all the major airlines.

When selling air fares, the systems are built around two key concepts:

  • A list of fares, at a fixed price point from City A to City B
  • Availability at each price point for every flight an airline operates

To visualise this a little better, let’s look a real world example from the GDS; a one-way, business class, fares from London to New York, travelling with British Airways.

Screen Shot 2017-10-16 at 17.02.33
A list of one way business class fares from London to New York

Note that the amounts listed are exclusive of taxes, fees and surcharges.

The list shows all of the fares that match my search criteria; there’s also the options to dive into a lot more information such as the rules and routing restrictions.

Next let’s look at the other key dimension of a GDS, the availability.  Again, for British Airways, travelling London to New York on 15th November 2017.

Screen Shot 2017-10-16 at 17.05.09
Availability display for BA flights from London to New York

The key column is “Available Classes” which is a collection of letters and numbers.  The number indicates how many seats at a particular level they are willing to sell, at this exact point in time.

The letters correspond to the first letter of the “Fare Basis” column in the fare listing in the first screenshot.

So putting the two together, if we wanted to travel on the cheapest one-way fare from London to New York (the INN8C1S6 fare), most of the flights are showing number I8 or I9, meaning there’s lots of availability.  However flights BA1516 and BA1506 are showing I0.  We’d need to buy a more expensive “R” fare if we wanted to go on those flights.

I go into more detail on these concepts in this blog post.

Why the need for NDC?

Particularly in the last five to ten years, airlines have started to unbundle their products, selling ancillary services along with the core flight itself.  These could be things like seat selections, extra bags, meals, in-flight wi-fi and so on.  Given the wide variety of options, and the fact that each airline has differentiated their own offerings and bundles, it’s simply not possible to offer this variety through the current, legacy GDS platforms.

These ancillary services are a huge revenue driver for airlines.  The low-cost carriers pioneered the concept of unbundling however in the last few years most carriers have adopted them with gusto.  They are very much here to stay, so the industry needs a way to sell them effectively.

NDC was announced back in 2012 and some five years on is still an active programme being run by IATA.  That fact that a non-government IT project is running over five years goes someway to explain how complex a business running an airline, and selling fares is.

The core concept of the NDC is to use XML (eXtensible Markup Language) and industry standard schemas to allow much more rich information to be stored in the GDS. It also allows airlines themselves to create standard APIs which developers can create apps against.

For example, the IAG group of airlines has a number of APIs available here, and the Lufthansa Group also has a set.

NDC enables other capabilities too, such as deeper personalisation and much greater control of pricing.

In summary, NDC is generally a good thing, albeit a massively complex global IT project that has to please many masters.

Implications of NDC

One of the interesting possibilities is the potential to do away with the concept of fixed fares combined with availability entirely.  Alex Cruz, the CEO of British Airways talked about this concept in his (in)famous interview with Skift back in January this year.

In it he talks about dynamically pricing airfares based on an airlines knowledge of the customer, rather than based on the current system I described earlier.

Rather than say “this is my list of price points” and “this is the level that I’m willing to sell my flights at”, there’s the potential to turn the concepts around and become completely customer-centric.

Rather than price an airfare based on what the airlines think it can sell for, why not try and maximise revenue by providing me a specific offer that it thinks I’ll be willing to pay?

Uber uses this technique very effectively with their upfront pricing in many US cities.  Uber now offers an upfront and fixed fare from A to B.  Previously when surge pricing was in effect, you had to accept the fact that the fare could be up to three times the “normal” rate (or pick an alternative).  With upfront pricing, Uber simply shows the total fare and you’re free to accept it or not.  This concept of dynamic pricing based on a number of factors is where the airline industry is headed.

However in the same article, Alex Cruz talks about the explosive growth of comparison tools to allow passengers to shop around and find the best deal.  He recognises the reality that there’s a hyper-competitive market out there and that there are some powerful tools available for consumers to find the best deal.

It strikes me as impossible to have truly dynamic pricing unless you offer incentives (i.e. discounts) for sharing personally identifiable information with an airline when asking for price quote.

For example, lets say I’m looking for a flight from London to Zurich on a Monday morning and coming back Thursday evening.  Given those flights times, I’m highly likely to be a fairly price insensitive business traveller.  A truly dynamic price quote from the NDC, if it has any sense, should provide me with a price that’s fairly high, but slightly lower than the competition so as to float to the top of the results. By adopting a NDC, the airline’s fare search engine can check out what the competition is offering, and potentially beat it, but only within a certain set of parameters specified by the business.

Death of loyalty?

In theory, this is all great, however the article which I referenced at the start is skeptical about this lack of transparency in NDC fare products, and I think that skepticism is well founded.  Airlines will do anything to improve their bottom lines – the basic economy product rolled out by the three major carriers in the US, is evidence of that. They created an even worse economy experience at the existing lowest price points, and then raised their normal economy fares which you now need to pick if you want a less awful journey.

Where I believe the death of loyalty may come from is that with the opaque and dynamic pricing opportunities available to airlines, there is nothing to stop them raising prices to frequent fliers. As soon as you ask for a flight quote, whilst “logged in” with your frequent flier number, it may offer a higher price that it thinks you will pay, rather than the best in the market.

Some customers may take the bait. Airlines only need a few customers to do that for them to increase their profits. However the more savvy ones will use air-fare meta search engines, and other tools to find better deals as they do today.  For example, booking as a new user, then adding in frequent flier details later on, after the booking has been completed and ticketed could be a simple workaround.

I suspect there will be an ongoing tit-for-tat game where airlines start to introduce this, and a lot of clever people attempt to reverse engineer the systems to understand what drives fare levels, and get around them.  The airlines will then adapt, introducing new techniques and the game will continue.

Where this will have most impact is on making changes to part flown bookings. Alex Cruz gave the example of a business traveller wanting to get on an earlier flight home from New York, with the system giving him an “offer” to do so, as opposed to the current transparent process of change fees and fare differences.

Airlines need to tread carefully to avoid alienating a significant portion of their customer base. However I think this is still some years off. Given the state of the NDC APIs that both IAG and Lufthansa Group have exposed so far, they’re not even at full feature parity with the existing GDS platforms.

Summary

NDC is slowly, very slowly happening. I don’t think the airlines have begun to scratch the surface of what they can do with it. In terms of the customer experience, being able to more accurately describe and price what you can buy from an airline is a good thing. The potential for dynamic pricing has some major down-sides though. It’ll be interesting to watch.

Leave a Reply