Blockchain & airline revenue
Airlines using blockchain with up-stream AI will achieve better results than those using AI on it's own
Think of blockchain and you might imagine hackers earning millions from strangely named cryptocurrencies. There is the original and turbulent Bitcoin, the supposedly stable Tether and even the wisecrack Dogecoin, which features an eponymous dog as it’s logo (dog e[lectronic] coin, geddit!). But cryptocurrency and blockchain are not the same.
Blockchains are built from a large number of computers working independently to solve common problems. In the case of a crypto like Bitcoin the problem in hand is demonstrating who owns each unit of money. The validated record of ownership thus created is called the distributed ledger and updated whenever a Bitcoin changes hands. This way, buyers and sellers both know the money they use is real.
Solving a mathematical problem to update the distributed ledger certifies the winning computer with “proof of work” and rewards the owner with a new Bitcoin, a process called mining. The computers managing the Bitcoin blockchain compete with each other to solve the challenging mathematics and for each transaction there can only be one winner.
So how is all this relevant to airlines?
One of the biggest talking points in airline commercial today is using artificial intelligence (AI) to figure out what products and services to bundle in an offer to show a travel shopper. But I think that airlines will get better results with blockchain. This article explains why.
AI sounds good but if delivered by the usual IT vendors it will be more of the same
IT vendors have always touted their whizzy algorithms as airline saviours, adding several percentage points to a carrier’s annual revenue and making all the difference in low-margin markets. Unfortunately airlines are not set up to make the most of the opportunities available (see article) so there is lots of talk but little action. Many chances to earn revenue are lost forever in the confusion.
As revenue management (RM) evolved (see article), each vendor’s boffins developed snazzy-sounding processes like “bayesian statistics”, “logistic methods” and “historical forecasts”. These estimated how many seats an airline would sell in the future and decide on that basis whether to accept a given price for a sale today.
The latest RM fashions are “shopping”, “rich content “and “offer management”. Shopping is the idea that people looking to buy travel products should be able to see exactly what they are buying, making them more likely to buy. Airline wonks analysing data on shopping can perceive trends before they are revealed through booking and ticket sales, and monetise accordingly.
The second fashion, rich content, is the inspiring pictures of a seat, meal or travel experience that support the shopping experience – the airline’s window on the high street. Airline fare publisher ATPCO bought Routehappy a few years ago to make sure that rich content and fares could be distributed hand-in-hand.
Final-fashion offer management is all about creating specific bundles of travel products like a flight with a seat reservation and a round of golf at the destination, plus of course free carriage of a golf bag. Airlines offering everything a passenger needs for their journey can earn commission on sale of non-air items, potentially collecting more than they ever would from tickets (see article), although there are unresolved fraud risks (see article).
Powering the latest trends is New Distribution Capability (NDC), a communications standard and pet project of industry trade association IATA. Unfortunately NDC has proved to be something of a MacGuffin as airlines have all rushed to be compliant with the technology without developing any relevant pricing and product strategies to actually monetise the opportunities.
NDC has promised much and cost billions but delivered little. Arguably airlines might have been better off each developing their own technology or engaging with startups. And as is was with NDC and the lost opportunities of traditional RM, so I fear it will be with AI.
AI alone is centralised high-cost, blockchain plus AI is de-centralised outsourced-cost
Airlines who go down the AI route can guarantee lots of meetings, high cost, implementation lag and delays. AI by it’s nature is a centralised technology that operates through command-and-control and only considers what it is shown by an operating airline. Blockchain on the other hand is de-centralised – how each computer solves a problem is up to it’s owner – and can include innovative data airlines might not otherwise access or understand.
So I believe that the best thing for airlines to do today is to outsource their revenue generation to a blockchain, with the blockchain’s individual members using AI not the airline. Just like Bitcoin blockchainers do with a distributed ledger today except the problem to be solved is airline revenue optimisation, not currency ownership. While I make no attempt here to resolve issues regarding the underlying mathematics, proof of work and incentive mechanics I propose the following:
· Airlines distribute either their schedule-capacity or their available aircraft & slots to a public blockchain that anyone can join
· Each individual blockchain member tries to find a solution that maximises the airline’s profit (net revenue less costs), using their proprietary advanced computation methods and likely including AI
· The airline implements the winning member’s strategy and is paid accordingly
· The airline does not need to pay members whose solutions are second-best.
Airlines using blockchain with up-stream AI will achieve better results than those using AI on it's own
This approach has three advantages:
· Because the blockchain is public, there is much more potential for innovation and market disruption than we see today
· There is a genuine opportunity for startups and scale-ups with better technology than today’s vendors to out-compete the incumbents, and little need for high-cost IATA-style management – airlines will be able to implement great ideas they would not have discovered with their own AI
· The airline outsources a large amount of opex and capex (potentially tens to hundreds of millions of Dollars a year for larger carriers), and only pays for what they actually use.
We should avoid an “IATA blockchain”
Since blockchains are public, how the work is done is up to the market. Airlines can adopt their own standards so there is no need to spend years collaborating with the IATA-vendor ring.
However there is a risk that IATA and the vendors will club together to try to promote an “IATA blockchain” that works on the blockchain method but is proprietary, high-cost, inefficient and suffers high barriers to entry.
Such a solution would be marketed as being in airlines’ best-interests, but in reality would probably only work well for a handful of the world’s largest airline groups.
What does the future look like for airlines and blockchain. Do we need to develop an airline coin? If you want to find out more, get in touch.
oliver AT ransonpricing DOT com