Amadeus explains its qualified support of IATA's Resolution 787 and talks NDC

Amadeus in its filing to DOT expressed “support for the objectives of NDC including airlines’ ability to differentiate and merchandize their offer”, and it has also expressed “support for attempts to standardize processes”.

We think our filing clearly, and in detail, clarifies our position.... We shared what we see are challenges with IATA several months ago, but have only publicly stated them now. Broadly speaking, we support a number of aspects of the creation of XML-based communications standards. But we also believe that many challenges have not yet been acknowledged or addressed. In our filing, we name four areas of concern, where we're looking for clarification. What we would love to see is that the standard is voluntary. Right now, we feel the wording of the document expresses a mandate. We'd like to see references around airline content ownership be removed because its unnecessary to move forward with a functional standard. We'd like to see the regulator step in and provide some guidelines around privacy. Because more data between travelers and airlines will be passing through the distribution system. We want to see a guarantee of backwards capability, which is what guarantees transparency.... We would like to see the resolution evolve to address these questions.... We optimistic that IATA wants to do this. For context, it's worth noting that Resolution 787 is a foundational document, and is only about ten pages long. Our internal XML API schema has documentation that's 200 pages long....

If the US regulator rejects Resolution 787, the challenges will not go away. I assume that, if it fails, the industry will start again and make a fresh attempt at a similar standards concept. Change must happen. It's important for the airlines to be able to differentiate their products in the marketplace. Amadeus already has its own building blocks toward this goal. We have an XML scheme in place in production. The way we connect with a number of local carriers today is not through EDIFACT but it's through an XML API schema. We are engaged in the distribution of ancillary product. We have 14 carriers in production in multiple markets. We're selling 50,000 to 70,000 ancillary services per month for our airline customers through the travel agency channel.... We are going in the same direction as Resolution 787, and we think the industry can pick up the pieces if this particular resolution fails. I'm not predicting it will fail, but just saying, if it did, industry change will still happen.

All stakeholders need to encourage adoption of the sale of ancillary services and differentiated airline products by travel agencies. But it's a very difficult process to manage. As an industry, we can get the technology in place, and airlines can supply the content, but some agents have been reluctant to embrace the change. Today, about 2% of bookings we facilitate in the travel agency channel include ancillary services. In comparison, about 8% to 10% of bookings in the direct channel include ancillary services. It's the same content, same services. There are reasons why we see much lower adoption rates. Offline travel agencies want to know, at the end of the day, 'What's in it for me?' They're worried about extra work, and potential extra costs in terms of connectivity for them. They wonder if they will make additional revenue in the booking process to justify the expense or labor to learn the new procedures. There seems to be a lack of willingness to have a discussion about this between airlines and travel agencies. For our own projects, we're trying hard to train and make it simple for users to learn the new procedures. But it's a huge challenge. As for the online travel agencies, you can see they control the evolution and development of their reservation systems. We are providing an API that provides ancillaries and differentiated air content, but it hasn't gotten as much usage by OTAs yet as we would like. Unless we have a tipping point of enough airline content, OTAs will be reluctant to integrate that content in their websites. They need critical mass. It may be that until about 40% of their bookings include ancillary services that they will be willing to adapt their websites to include our content. Even if the NDC gets traction on the airline side, you'll still see obstacles to adoption.

The biggest challenge, from our perspective, is about shopping and scalability. Today we are in a world where the requirements coming from the traveler are going through the roof. Travelers want millisecond response times, they want to search on maybe 100 dates between an origin and destination, and so forth. So the transaction resources required are increasing. Capacity planning is critical. One way to respond is caching. Amadeus is providing a variation on caching. We're pre-populating the price and seat availability to guarantee rapid response times. Online travel agencies rely on our cache and then add other sources of data to deliver their own results to the traveler. That's taking place today. In a way, one of the challenges we can see with a dynamic pricing approach is that you're going to have to choose between caching and personalizing the product. It's going to be tough to reconcile both. If to get price and availability on an offer for a travel agency, you as an aggregator need to automatically query an airline's system to get a quote out, it will become much more complex and much more intensive in terms of resource consumption. You've got scalability and bandwidth issues first. You've got a cost issue second because an increase in transaction resources will translate into costs for many parties. And that will all need to be sorted out. This is a nuanced point. Today, even though we power many queries by cache, which is very efficient, we are dynamically pulling from the airlines for availability. What we're doing, even though the fares the travel agencies retrieve in the system are managed in a separate cached system, we are checking the availability in real time. Which is close to an NDC-style concept. We see the number of these transactions are rising, and it's mostly driven by the online channel. So already today, without the NDC, because for online travel agencies the look-to-book ratio is so low -- it's not rare to see a look-to-book ratio of between 100 and 1,000 for an OTA -- and because we don't want to cache everything and we do want as much real-time as possible, already today it becomes challenging for the model to work. If you push to the limit and say you have to do dynamic packaging for everybody, it is going to be a big strain on the system. And that's the constraint you're going to see. It's an important technical aspect.

Today 15% to 20% of airline tickets sold through travel agents are interline, a concept which your readers are familiar with. Basically, airlines agree to jointly market an OND (an origin and destination), and one carrier will get the money on the ticket and split it with their partner. Interlining is really important for a network carrier, because it really expands their reach. The way it works today, they're communicating the terms of the interline contract through the global distribution system. We build the offer in the system for flying between A and B on two airlines, and the validating guy is going to be the GDS. That's one of the benefits of using a GDS. We're uniquely placed to construct the interline offer in our system. Tomorrow, if you follow the NDC approach, we might have to break that logic. You would need one airline to actually provide all the elements of the offer, such as price and seat availability, for both themselves with all of their potential interline partners, to the distribution system. As an airline, you would need connectivity with maybe 20 or 40 airline systems, which is not the case today.

Looking beyond a standards change for messaging, there are the future questions of the NDC and how airline content should be distributed more dynamically. I can understand how a consumer advocacy group might look at this from a certain angle. As both an IT player and a distribution provider, the way we look at this issue is about enabling comparison. Transparency is only possible if content can be compared. Airlines should have a vested interest in making sure their content is relevant. I don't think most airlines want to withdraw content because that would mean moving out of the market. That's why we want to see references discouraging backwards compatibility be removed. Wherever the content comes from, NDC or not NDC, there must be comparability in the content because that is what is going to guarantee transparency for the travel agency and the consumer. Backwards compatability will be beneficial to the airlines themselves. If some airlines stop providing some content to the market or if their content via the NDC cannot be compared with content coming in EDIFACT or via APTCO filed by other airlines, then they won't achieve their goal. Their content won't be compared. The consumer will want to be able to compare, and airlines need to respond to that consumer behavior by providing relevant content.