skip to Main Content

Using Data to join the Dots in a Brave NDC World

NDC world data analytics

N-D-C. There I have used the magic three letters on line one. Three letters which will soon be consumed and superseded by IATA’s One Order but that story is for another day.

Talking Points

There is a lot of talk out there, by that I mean discussions, presentations and even on occasions heated debates about the changing world of airline product distribution, NDC will likely change distribution bringing greater control, dramatically shortened time to market for products and perhaps most importantly visibility of search traffic not just the recent booking history of traditional channels.

It’s not just NDC conversations, there are other major discussions:

  • Ancillary Revenues will be above $60bn by the end of 2016. These aren’t just relabelled unbundled luggage revenues but genuine add to the bottom line dollars which are usually on a much high margin basis than the seat itself.
  • Retailing. You often hear from industry commentators that airlines need to reinvent themselves as the next Amazon. Probably not but they certainly need to get merchandising in a more cohesive way. They need to grasp the opportunity of product differentiation in the indirect channel using features/benefits and not just price – something that NDC will enable.
  • The Offer is the new biggest asset for the airline. So the shiny fleet of $100m aircraft play second fiddle to the hundreds of millions of shopping responses (offers) that will be made to prospective travellers via various channels every day of the year.

The 21st century customer is thoroughly, if not overly, connected to everything in and out of sight. He or she is probably more au fait with the competitors’ offerings than the airline he is currently conversing with. Failing to put him first and not understand him will result in a very short relationship. The real value with most customers is in their LTV or life time value.

Joining the Dots

These are all great thoughts but how we do really join the dots to get a meaningful picture that an airline can embrace as a vision?

I believe the most significant part to the answer is data and I’ll come back to that idea. But it’s also about adding new technology such as extending the PSS infrastructure to support NDC APIs at scale, process changes such as directly managing offers and their pricing very dynamically rather than via cumbersome traditional filing, it’s about hiring people who can merchandise and it’s about skill sets such as the ability to analyse data and drive decisions.

So back to the data side of things. The only way an airline can successfully engage with the modern traveller in a customer centric manner is to start by understanding him. There are millions of ‘hims’ so that means using data to identify segmentations of like-minded individuals and help formulate offers that can road tested and revised continuously until they are the best that they can be.

Let’s assume we have a decent data analytics system which can silently and seamlessly capture and reverse engineer all my NDC API traffic so I can get a clear view across six months’ worth of searches or bookings and ancillary transactions from all my OTA partners and NDC aggregators. The system captures all traffic for all flights, but I want to just look at the events leading up to departure of a single flight from London Heathrow to New York JFK on 10th October. The filtered chart might look something like the chart below.

Customer Buying Behaviour: Transactions for LHR to JFK Departing 10th October

Join the Dots

Excusing the pun, this data chart is truly a mine of information.

We can readily see where the peaks in shopping (search) occur and trailing behind at a lower level the bookings (PNRs) and then finally when ancillary revenues (EMDs) get purchased – including a second peak just before travel date – part of the so called second wallet where post booking, travellers make last minute purchases of anything from extra hold suitcases to rental cars. We can calculate the average Look Ahead (days between search and departure date) and average Book Ahead (days between book and departure date) with the gap being decision making or approval seeking. Approval could of course be from the line manager in the context of business travel or the spouse in a leisure situation.

If we used additional search parameters in our analysis we might look at the day of week, duration of stay and other factors to segment this into an (estimated) business travel segment. We might see sub peaks in the shopping/booking consistent with for example long term conference attendance or short notice relatively unplanned business meetings. The business approval process might be different and the offers needed may also be significantly different – one is actually quite fixed and the other may benefit from flexibility.

If we decided this was a leisure trip and perhaps it is spousal approval dependant and if the peak booking point was near a very modest but timely offer might just secure the PNR.

The idea here to just to whet the appetite on the sort of analysis that good quality data can support and the type of hypotheses it can confirm or deny.

Clearly putting the right NDC related technology and processes in place affords the airline the opportunity to start truly driving better conversion and ancillary revenues through the indirect channel in much the same way as some airlines may already do on their brand websites.

That’s said, the best merchandising rules engines, booking platforms and PSSs alone will not alone deliver the levels of revenue required. It is the data that joins the dots.

The bottom line is that airlines may need to get merchandising but to be effective they really need to get analysing.

Footnote

NDC stands for New Distribution Capability, the XML standard for airline distribution being proposed by IATA. Triometric has previously published articles on the general need for NDC including a background whitepaper and a more detailed set of example BI charts including more complex analysis taken from NDC API traffic.

Back To Top