Revenue Management Isn’t Cancelled

We’re two months into 2021 and despite the hope that a new year would result in a new outlook for the pandemic, we’re all still seeing the deep impact across the industry. Despite the negative near-term forecasts, I believe that we’re all starting to see that glimmer of light at the end of the tunnel as the vaccine rollout gets fully underway. I’m optimistic for the future and what that will mean for the airline industry. I also think there’s an opportunity to learn from what we saw in 2020 and how we can apply new strategies in 2021, especially from the lens of revenue management.

If there’s one thing that I’ve repeated over and over in the last 10 months, it’s that Revenue Management isn’t irrelevant and isn’t canceled. This has been a popular sentiment expressed in mainstream media, as well as throughout our industry. I understand where the idea comes from. If you’re using history to predict the future, it is certainly going to be challenging. But the principles of Revenue Management are still valid. Practitioners are still trying to determine the right product at the right price, at the right time, for the right passenger. With so many moving parts and complications like capacity, scheduling, and competitor movements in the industry, a robust and powerful revenue management system is still necessary. The key is having the right strategy for your RM department and the right system to do that.
 

Can You Trust Forecasting?

This really is the crux of the feedback that we’re hearing and seeing around the industry. The assumption here is that what happened in the past isn’t relevant anymore because the future has changed so dramatically. There is some truth to that, especially if you were using a simple year-over-year model. But PROS doesn’t take this approach, so I want to share more details around it. We use a proprietary Bayesian forecasting methodology developed specifically for forecasting passenger demand. The Bayesian part of the algorithm is simply describing the underlying statistical model. Many AI and machine learning models use Bayesian as their underlying framework. But what can really set this model apart and drive value for airlines, even in extreme volatility is:

  • Establishing correlation between different parts of the booking curve
  • Using booking data from flights yet to depart and weighing more recent observations stronger
  • Automatically capturing the seasonality of the curve

Let’s talk about each of these areas in the context of the "guillotine".
 

What Really Shapes the Forecast?

The picture above really explains each of the three components I mentioned above so well. We affectionately call this the guillotine. You can likely tell why with the shape that it makes. The good news is that it isn’t nearly as violent as it sounds. Down the left-hand side are the departure dates.  Across the top are the days prior to departure. Each cell represents the observations of bookings for the entity we’re ultimately going to forecast. 

First, you can see that we use cells that are below today’s date. We’re already taking into account bookings for those dates and using them to adjust the forecast. This helps us drive the seasonal shape of the bookings. Additionally, we are sharing information across the columns. We understand what is happening in different parts of the booking curve and use that information to adjust other parts. Lastly, while all of the data from the history is used to feed these calculations, we’re also able to weigh the more recent data to account for the fact that the more recent observations are more reflective of the new reality of the market. These three key components balance together to create a forecast that uses the good parts of the history with the most recent information to generate a forecast.
 

How Can Airlines Fine-Tune this Forecast?

On top of this, each of the components mentioned above have configurations that drive how strongly they are considered. These configurations allow airlines to determine should more information be shared across the booking curve, or less.  What sort of cycles would be observed in the seasonality?  How aggressively should the recent observations should be weighed? These are all critical questions in the COVID scenario. Given the volatility of this time, the airline can adjust these settings to be most appropriate for the conditions they’re observing.

Now you might be wondering, “how do I know how to set those?” It’s a great question and one that is exactly why we formed the COVID-19 Task Force. The carriers participating in the Task Force have shared their data with PROS, which was then combined with third-party data like infection rates and country closures.  With this data, studies were conducted that output recommended configuration changes based on the unique characteristics of their markets. 

The methodology and configurations together result in a forecast that is robust and in line with the current conditions in the market. The forecast continues to learn and adapt as more information is observed. This is the power of the PROS forecasting methodology.

But that isn’t where it stops - revenue management is far from being canceled and the tools and components of PROS Revenue Management are critical for airlines today. I cover a lot more of this in a presentation I did for Aviation Festival London this past fall.


Stay tuned for my next post, where I go in-depth on the fare valuation and optimizer components. As always, if you have questions or if you would like to dig deeper on any of these RM topics, please reach out to me.

About the Author

Justin Jander

Justin Jander is a Director of Product Management, focusing on the Revenue Management products at PROS. Justin has been with PROS for 11 years, all within the Product Management group, focusing on the travel products. During that time, he has overseen the continuous improvement of the PROS revenue management products. In order to understand the needs of the always-changing industry, he has worked with airlines across the world, which allows him to understand the business problem and translate that into features that can improve the RM system. Justin earned a Bachelor of Science degree in Mathematics from Stephen F. Austin State University and a Master of Science degree in Statistical Science from Southern Methodist University.

More Content by Justin Jander
Previous Article
Winc Australia Smooths Sales Process with Pricing Data
Winc Australia Smooths Sales Process with Pricing Data

Hear how office supply distributor Winc Australia optimized price guidance for its sales teams to automate ...

Next Article
Smart CPQ: Fully Integrated with Dynamics 365
Smart CPQ: Fully Integrated with Dynamics 365

Learn how Smart CPQ integrated with Dynamics 365 helps shape winning deals and optimize your pricing faster...