More on Retailing: Segmentation, relevance & data

by Alexander v. Bernstorff

 

PERSONALISATION has become one of the buzzwords in aviation, at the latest since IATA launched its NDC Program to enable the distribution of offers that are self-assembled and personalised, instead of pre-configured and assembled by 3rd parties. Two interesting white papers I found recently suggest that it takes a few more words than usual to deal with this subject.

Segmentation and relevance

In their white paper called “Personalization, Value and Relevance: A Pragmatic Approach”, the think-tank of NRF initially asks the question “What if relevance is the goal, and personalization merely a tactic to get there?”, assuming that retailers “don’t fully understand what personalization is, or how it differs from segmentation”. Aside from the ado about personalisation in our industry, even very large airlines with big capacities of business experts are struggling with making up a customer segmentation that reflects current consumer habits and expectations. Consequently, airlines suffer from a shortage in self-controlled distribution channels, designed around the customer, that strange animal.

Control of distribution channels is one of the most powerful drivers of profits in any industry, and this is especially the case in the airline business.”, says Stephen Shaw in his standard “Airline Marketing and Management”. Accordingly, NRF concludes that for positive examples, ease of use (e.g. within a mobile shopping site) might actually have something to do with an understanding of customer needs and desires.

It seems that the combination of a proper customer segmentation, alongside distribution channels these customer segments wish to use, is an important step towards more relevance. With an NDC API, airlines can faster and more easily design and develop distribution channels than before.

The role of AI

Not forgetting: AI (Artificial Intelligence) is right here, in the middle of our industry – in experimentation-mode, however. We can see various examples where AI in its different forms can be used to help interpreting data better and come up with practical solutions. On the one hand, it scares us because it is suggesting to be able to make better decisions than we can today, based on our long experience in what we do. On the other hand, it may provide useful information to come up with a better customer segmentation, more relevant offers and less friction for the consumer. The challenge is to stop thinking and working in silos, but to exact open interfaces for our brains as well as our technology in order to be able to incorporate insights delivered by AI. For example, an AI-based chatbot can deliver robust information about the problems the sale of a particular service offering creates during fulfilment. With a flexible retail engine, the service could be de-activated with a click, then adjusted to address the problems reported by AI, and re-activated in a better way. At a later point-in-time, AI could automatically de-activate that service and propose changes to the responsible department within the airline – leading to more relevant offers.

Small data analytics

Adding to segmentation and relevance, another white paper, based on research by SAS (the analytics experts, not the airline) and Roland Berger, finds a certain degree of negligence when it comes to utilising data that is easily available within their own companies. The university of Potsdam (near Berlin, Germany) found, that almost 90% of businesses analysed will use less than 50% of data already available to them.

Forget about BIG DATA and replace that with “smart data” and there you go: it seems to be smart to spend some time on connecting data-pools already available and making sure such data, properly analysed and interpreted, can actually support decisions (We do want to go from experience-based to data-based decisions, don’t we?).

Now, what’s this data thing about? It’s a long way from “people who bought this also bought” to “based on your preferences and history, here’s what you should buy next”. Today, airlines have little knowledge about what people shop for, what is offered to them and what they finally purchase (if they purchase). Imagine why clothing companies put their own salesclerks into their shops-in-shop within large department stores. They need to know as exactly in indirect sales as for their own shop who is shopping, what people shop for, what’s on offer and what is finally being purchased.
To make a long story short: Retailers (and airlines) should own their data and properly analyse it in order to come up with a customer segmentation – and ultimately the right distribution channels. Interestingly, SAS and Roland Berger found that retailers are pretty advanced in this area – another argument also for airlines to look closer at what happens in retail.

Guess what: NDC has the potential to store and analyse shopping requests, offers and orders!

Fact and fancy 1: Shopping for flights is price-driven

If someone talks about price being the only, or by far most important buying decision: Pretend you didn’t hear that! Probably that person works for a price comparison or other shopping enterprise that is not capable of incorporating aspects, other than price, into their shopping flow. Think of your own shopping behaviour. VentureBeat research found that the vast majority of consumers (9 out of 10) are willing to pay a premium, if they only got what they wanted. But only 1 out of 100 feel their expectations are being met. Disappointing, huh?

As also outlined by IATA’s latest global passenger survey, airline customers are indeed willing to pay a premium (out of their first or second wallet) for add-ons that make their journey more pleasurable. Airlines should be able to package, bundle, offer, sell etc. many different products and services, produced and fulfilled by themselves or by third parties, to be able to make more relevant offers to their customers. And that’s a fact, and a modern NDC-based retail engine can help.

Fact and fancy 2: My brand is my castle

Bad news: According to SAS and Roland Berger, the times when companies, through advertising, defined how and where customers had to interact with their brand, are over. Customer sovereignty means that two thirds of all interactions are initiated by the customer, rather than by a company’s marketing department. And the customer decides which channel to use. Many enterprises seem not to be prepared.

Let’s be honest: Most airlines did not manage to pull the majority of their customers onto their own website. When talking to airlines, we can still see a preference to invest heavily into their own shop, where they can control their brand, instead of investing into capabilities that allow them to push their brand to where the customer prefers to be.

That does not only increase the relevance of a proper omni-channel strategy (not to be mixed up with an all-offers-in-all-channels strategy, which is fundamentally counterproductive), but it also makes clear what the value of rich content distribution capabilities is. By the way, this is impossible in the main distribution channel for airline products today, but nicely possible in NDC.

Will it work?

Risk aversion, something too often found in airline commercial management, usually provides for complex and time-consuming specifications and RFP processes when it comes to new IT systems. At the time, a product has ultimately been chosen and implemented (including massive training, thousands of bug-fixes and so on), that system or software may already be outdated (remember, it’s been only 10 years we have smartphones). But all good, since the airline expects this system or software to solve their problems for the next 15 or 20 years.

Unfortunately, we need to be AGILE. Both white papers strongly suggest to “create a culture of experimentation”, based on an agile culture and equally agile IT systems, also taking MVPs into consideration. 80% of enterprises surveyed feel that agility has become essential for their survival, whereas 60% feel that they are not prepared.

Believe me: the Swiss Army knife does not exist. It is essential to build an architecture that allows for true BEST-OF-NEED and enables testing, experimenting and to reduce the cost of failure. Remember, Amazon is doing up to 1.000 deployments a day.

The answer is, thus, yes – it will work.

Talk to us

We are keen to discussing buzzwords, agility, retailing and other interesting topics with you, and also show you our NDC & ONE Order MVP, an offer- & order-management system designed from scratch to enable true, simplified airline retailing (based, however, on 20 years of experience serving top-tier customers like TUI and Swiss International Air Lines).

Schreibe einen Kommentar

Deine E-Mail-Adresse wird nicht veröffentlicht. Erforderliche Felder sind mit * markiert.