Marios Prokopiou co-wrote this blog post with Evan Gray.
For more ZS insights on the impact of COVID-19, visit zs.com/COVID19.
The widespread lockdowns in response to the threat of COVID-19 have forced the travel industry into uncharted territory. But as peaks from pre-lockdown start to pass, and as local and national governments consider long-term solutions to manage the spread, demand will begin to gradually recover, and airlines will incrementally restore capacity. During this likely recovery period revenue management teams will need to fundamentally change their short-term strategies, models and metrics.
With travel restrictions changing by the day, the industry needs to take a more agile approach for creating strategies and measuring performance. For revenue management teams, it’s not enough to just adjust demand forecasts. Existing strategies for incremental revenue should be put on hold in favor of short-term, tactical influences with easy-to-measure impact. Establish new ways to measure incremental revenue gain. Year-over-year yield comparisons will be less useful than RASK comparisons and month-over-month market share changes.
Before starting to clear out existing influences and optimization rules in your system, consider refreshing your revenue management system sandbox environment. Use this to study the impact of removing existing forecast influences and create a plan for transitioning them back post-recovery. The “why” behind these influences may still be valid post-COVID-19.
The evolving demand patterns during the recovery period will challenge existing region-based team strategies. As different regions are impacted by the virus at different times, each will be on its own recovery timeline. Make sure your team is aware each of market’s phase of recovery and shares success and failures from these to improve your future recovery strategies. For example, once the time scale of demand recovery for a market has been observed, demand ramp-up influences with that time scale can be created for those markets still early in their recovery. Also, non-focus markets, if recovering earlier, can offer valuable opportunities to test strategies before your focus markets recover.
If your team uses QSI or fair market share models, many of their underlying assumptions may no longer be valid. These models are typically trained on historical periods with healthy demand and full capacity, which no longer resemble market conditions. Outlier low-demand periods, previously excluded from a model’s training set, could now be valuable. As post-lockdown data becomes available, shift weighting to recent periods. For QSI models that compare the attractiveness of different itineraries based on features such as departure time and number of stops, the relative importance of different features will likely change. Forecasts should be updated more frequently as carriers continue to adjust schedules and capacity.
A common method of forecasting in many airlines is trend-based, using historical sales as a guide and historical analogs to predict passenger demand. However, trend and analog forecasts are likely to be insufficient in light of COVID-19; recent global recessions or past pandemics do not speak to the magnitude of disruption. To better predict new sales, airlines should include some element of demand-based forecasting and widen the variance on probabilistic scenarios used to simulate forecast outcomes.
Recovering from the COVID-19 disruption will push revenue management teams to increase their agility, adapt to new forecasting methods and KPIs and re-examine criteria for strategic success. While changing foundational aspects of your team’s approach can be daunting, the lessons learned can lay a foundation for a more effective team beyond the recovery period.