Ricardo Pilon is an experienced aviation expert building the bridge between strategic business management, organization design, and the application of deep technology and applied psychology.

Subscribe to his Airline Behavioral Economics newsletter which now has over 33,000 subscribers gaining valuable thought-leadership insights on servicing retail demand with deep tech using applied psychology. 

There has always been plenty of debate about whether redlining a car’s internal combustion engine is good for your car or not. Petrol heads like to say it’s good for engines, but many think that screaming and squeaky engines sound like they’re about to explode.

Redlining means letting your engine rev until it hits the limiter. An engine needs to rev to a certain point in order to produce its maximum power. That’s why you see technical specifications that say “550hp at 6500rpm”.

So, engines are designed to spin to redline whenever maximum power is required. In the above example, 550hp is produced at 6500rpm. And in fact, most engine redlines are set a little higher than where maximum power is developed.

"The scope of today’s revenue management is very much limited in its commercial optimization capabilities, despite some recent incremental improvements in AI-assisted forecasting, and both dynamic and continuous pricing."

Redlining your engine every now and then has several advantages. It helps keep moving parts in good running order while bringing out any deficiencies that may not be obvious at lower running speeds. Like partly clogged fuel injector nozzles and worn spark plugs. But there’s no need to worry.

Redlining will not damage an engine or cause it to explode, no matter how cruelly you treat it (but don’t do it when the engine’s still cold).

An engine will rev and perform as it’s built. Normally.

It will do what it’s designed to do, but overall car performance and handling depend on many other factors that the engine cannot handle. Or consider. It must come from other realities it’s not connected to.

The same applies to airline and hotel revenue management. There, it’s actually pumping the brakes! Revving it won’t do much more.

The scope of today’s revenue management is very much limited in its commercial optimization capabilities, despite some recent incremental improvements in AI-assisted forecasting, and both dynamic and continuous pricing.

But it’s marginally incremental.

One of the organizational and psychological problems is that the term revenue management refers to a department and not a (potentially new) function. This function, which I like to call “Commercial & Customer Revenue Optimization” cuts across silos and units in a newly formed organization. It has B2C and B2B components plus aircraft cabin applications and we include physical and digital panels monetizing impressions.

In this first part, I’m only addressing the B2C component.

I recommend deploying more holistic enterprise methods like '“EMCRs - Expected Marginal Customer Revenue Streams”, optimized with new Customer Balance Score Cards (CBSC).

Read on why revenue management’s future is outside the revenue management department.

Gatekeepers (Inventory)

Revenue management was designed by smart and slightly geeky introverted people as a Blackbox calculator. In my early career days, we built the formulas using Expected Marginal Seat Revenue (EMSRb) models with inventory rules based on differential pricing for Roots Air 23 years ago. All of us worked for other airlines before, including KLM and Air Canada, and now the vendor’s software solutions.

"RM is a gatekeeper that controls which seats are allowed to be sold at which price points, and through which PoS."

I often say that revenue management (RM) does not sell anything. They don’t. RM doesn’t interact with customers. They do this only through Sales, for corporate sales programs, and through the Group desk. RM does not know the individuals that travel, nor their full travel context or preferences. Segmentation is at a very (anonymous) high level.

That’s how P&RM can be ‘insulting’ to many when they shop.

(Source: unhhockeyblog.blogspot.com)

So, RM is a gatekeeper that controls which seats are allowed to be sold at which price points, and through which PoS (Point of Sale, like channel, agency, physical location). Traditionally, those gates were for the ‘trade’ (the intermediaries) who use GDS’.

Nowadays, the gates also depend on (a) pax shopping, [‘some’ of the] (b) older history, and (c) the requests actually coming in for various fare products and multiple channels. The software RM people get used to does not reveal any identities, although PNR-based forecasting has been possible since 2002. But it’s all about seats and some ancillary (like seat baggage, seat selection, insurance, and others).

That’s their role. That is the scope. It’s a small scope.

* * * * *

Inside-Out (>production>assembly>)

The gatekeepers essentially apply rules to the inventory that is created by the schedule. This production of seats is sold through intermediaries (agencies) or direct online booking engines (incl. mobile). Much of that depends on the structure within the distribution channel (e.g. GDS) – the infamous A-Z buckets). We created all sorts of interesting nesting structures, but RM is still an extension of production (fleet > network > schedule> Origin-Destination markets>segments>directionality).

It’s very linear. There’s my problem.

While the New Distribution Capability (NDC) allows richer content (pictures, reels, dynamic pricing, servicing) it is not retailing. Especially when it’s about using NDC interfaces to the TMCs or other travel agents. That’s wholesale.

Or when “content” is offered directly to the customer. We are touting fancy Offer & Order (O&O) management on top of PSS’, but it still reflects the stackable approach of ancillary products & services, especially the path.

That’s no different than what fundamentally is upselling and cross-selling.

There’s my other problem. It controls directions.

(Source: Robotics Business Review)

Outside-In (retailing<state of mind<customer)

The reason why the future of revenue management is outside the RM department is that we need consistent customer experiences. And that involves feelings of what is considered fair, just, and reasonable, in each customer’s context and past.

It requires multiple frames of reference that have little to do with travel or destinations.

And, that includes emotions around all touch points. The customer is the starting point, not the OD or seat. And differential pricing and timing have to make sense or at least be explained better.

We have to start working from the outside-in and embrace real retailing. With effective retailing, the product is not necessarily the central point. The environment is. And what it is compared to outside travel. Plus, how we encourage foot traffic. It depends on people’s emotional responses (see article) to all types of content and how it’s presented.

"The reason why the future of revenue management is outside the RM department is that we need consistent customer experiences."

So, digital marketing tech (MarTech) and SalesTech underpinned by a combination of AI technologies have the necessary capabilities to interact with real customers and create the atmosphere that is conducive to shopping across multiple shelves, in a way the customer chooses. Like how they find their way around Macy’s or Harrods.  

It also means that as customers consider purchases, the full package is constantly (dynamically) rebundled, re-offered, and other content dynamically procured that now fits the full context. Even if it’s through a ‘competitor’.

The point is trust and stickiness. Long term. It’s loosely related to Harrah’s Entertainment’s tactics in casino accommodation (share of wallet) but further modernized.

It should follow at least these principles we can adopt in aviation and travel:

  • The digital store has the appeal of and aligns with the customers social and diversity values.
  • It encourages social proof based on the passenger’s choice of community (like Tripadvisor), and even recommends competitive offers for full transparency to built trust and loyalty.
  • It weighs sub-optimal offers against premium margins that is overall goal-centric (deep learning can assist).
  • It allows customers to define their needs themselves, like arrival-based shopping or using free-style words to describe the purpose (like Air Transat advertising on how you [want to] feel when you’re back home).
  • Enable clients to set initial budgets and trade-offs they see (air, land, experiences, and quality vs quantity).
  • Allow fast-track (flight only) options to check out quickly (don’t be afraid to lose out!), which can always be expanded.
  • Clients can store digitally assisted playlists and interests with AI based on their parameters (convolution algorithms).
  • It must be tested to function well onboard in-flight, where the passenger is static but cabin fixtures and other platforms can promote and transact relevant offers.
  • (Truly) Dynamic stacks are built with the customer based on the margins of each offering and the desired loyalty level supported by real-time procurement.

* * * * *


Because of the need to focus on individual customers and uniquely detailed contexts, this level of personalized retailing experience can only be provided by retail, customer and user-experience specialists supported by behavioral psychologists.

This talent exists or can be obtained across the airline organization (see article), but it does not fit in today’s RM departments. It also doesn’t fit in the mentality, the ingrained beliefs and psychological barriers of what “revenue management” is (see article).

RM is what it is because it was defined by RM-ers and vendors, not CCOs.

That must change.

And the best way to do it is to use a new name & org that captures what commercial optimization is [bigger picture], and that RM is only one component that contributes some insights.

"RM is what it is because it was defined by RM-ers and vendors, not CCOs. That must change."

RM is also biased by the way it looks at the discipline using the current generation of software. It does not ‘manage revenue’ nor optimize it because other factors play a big role (like personal views on competition, randomly-set market share targets on specific routes and markets), and the sub-optimization is not captured by any system.

If I was overlooking the entire commercial space now, I would create a ‘Commercial & Customer Revenue Optimization” organization within the CCO’s realm.

It would be:

  • Driven by the loyalty program, focusing on individuals.
  • Looking at all revenue streams and make a balance between past, current and predictive spend, before making custom, optimized offers in real time.
  • Allowing the loyalty program to create consistent retail procurement-offers.
  • Building the bridge to digital marketing and sales to identify and serve prioritized customers and markets.
  • Using traffic forecasts from the conventional RM unit.
  • Allowing the RM team to execute business rules that protect and serve loyal customers, as per loyalty’s leadership.
  • Using enterprise AI to build the links between the functions to create consistent, harmonized end results defined in customer metrics.

Here’s a higher-level overview combining functions on platforms (below). All conventional BUs are service providers, governed by CCRO on the social commerce marketplace.

* * * * *

Techy Stuff & Vendors

Clearly, the techy stuff is tricky. Because it doesn’t yet exist. The software packages are built to support the very silos they helped create. But there is an opportunity for (1) airlines, (2) software product, and (3) software services companies to move up to the enterprise level that will support the CXO team better using future tech generations including deep learning.

The first steps are being made, and they actually started in the cabin. Because it must come together properly in the captive state differently that in-seat order management systems. That’s just incremental.

Like today’s customer, their online behaviors, and the multiple channels they use simultaneously, there is nothing linear about it. It’s omni-directionalomni-functional, and must be consistent no matter how they contact the airline or hotel.

This is especially the case for how they service customers post-sale. That can only be achieved with a tighter organization that brings fleet, network, schedule, marketing, sales, distribution, pricing, RM, together on a layer that is supported by Enterprise AI.

The end models are goal-centric (customer goals and profitability). Without blatant conflicts.

So, that’s The Big Picture, the goal, delivered in steps (see video). But not out of RM. They don’t sell and no matter how much they rev themselves and the RMS’, they’re always hitting their commercial limiter.

We need a better and higher limiter.