It generally happens at least once a year: You go to the doctor, and the nurse initiates procedures to measure your blood pressure, level of oxygen, “good” and “bad” cholesterol, sugar level and any number of other key indicators related directly to your fundamental health status.
If you currently maintain good health, the annual testing procedures will not be an issue.
If not, however, the doctor will likely draw up a list of recommendations that may include — but are not necessarily limited to — a more intense level of exercise, healthier eating (in other words, you should probably not be eating many of the things you like) or taking specified medications.
Airlines, very much like people, go through health checks annually and monthly. Even more meticulously, they measure some essential factors weekly and daily.
Similar to a doctor, an airline manager pores over a series of key indicators to measure the health of the operation. Most importantly, he tries to interpret precisely how these measurements relate to the airline’s overall strategy.
In studying the details of this airline examination, we need to understand exactly why operational key performance indicators (KPIs) are important, where the majority of data points come from and are stored, how to set up an airline’s organizational goals, and how airline executives can measure the overall health and robustness of their airline’s network.
Linking Strategy With Executives
The main reasons to track operational KPIs is to understand how they perform with regard to an airline’s strategy as well as to compare (or benchmark) against competitors and track operational efficiency.
KPIs also provide input for continuous improvement processes performed by the airline and allow the early detection of any operational problems.
Finally, operational KPIs reflect the robustness of the airline’s network.
When an airline’s chief executive officer gathers his or her team to define next year’s plan, one of the key components is the level of service the airline intends to provide customers.
Because an airline constitutes a service business, executives always focus on on-time performance as one of the paramount customer-service indicators.
For example, when Gordon Bethune and Greg Brenneman turned the Continental Airlines organization around, they did it based on a set of strategic goals and a set of KPIs to track the progress of the goals. The plan included four pillars with catchy names so everyone across the organization could easily remember them.
The focus on the operation was considered so important — organization wide — that the marketing plan (the organizational cornerstone) was called: “Make Reliability a Reality.” The overarching strategic goal was to improve the product — to become an airline of preference.
To achieve this ultimate goal, Bethune and Brenneman tracked other airlines in a series of key indicators. The idea was that Continental should perform in at least the top 50 percent of the industry’s KPIs that are closely followed as general industry indicators by the U.S. Department of Transportation.
This approach is not new. In fact, it is supported by many scholarly studies and writings. Robert S. Kaplan and David P. Norton, authors of The Execution Premium, have written extensively about how companies should track the execution of their strategy using a balance-scorecard approach.
It is not our purpose to delve deeply into these methodologies. However, any C-level leader at the airline needs to understand exactly which KPIs are critical in measuring the service provided to the customer and how those KPIs compare to the airline’s competition.
This knowledge could substantially support the management team in developing a plan or a set of goals that will underpin the considerable organizational strength that is required to bring about serious change.
This concept (as demonstrated in Figure 1) has very understandable tenets, such as the logical assumption that passengers want and expect to arrive on time based on the airline’s schedule, which passengers regard as a promise. This concept links to two primary metrics: on-time performance and completion factor.
KPIs For A Healthy Airline
Naturally, there must be alignment between the strategy and each KPI, including internal alignment as well as external benchmarking, both of which are required to move the company in the desired direction. The gaps will drive the analysis to understand the required changes. A key set of KPIs — departure performance, block performance, arrival performance and completion factor — must be tracked by an airline’s senior management to manage change and ensure a healthier overall airline operation.
As shown in Figure 1, alignment between the strategy and each of the KPIs is required. It also shows that, in addition to achieving internal alignment, to move the company in the desired direction, external benchmarks must also be part of the analysis.
The gaps essentially drive the analysis so required changes will then be fully understood by those managers and team members entrusted to implement them.
In addition, Figure 1 reveals a KPI set that must be followed closely by C-level and senior management. This KPI set encompasses departure performance, block performance, arrival performance and completion factor.
Other essential KPIs are the mishandled-bag ratio, turnaround performance (or “turn” performance), passenger complaints and passengers misconnecting.
Just to be clear, there are many more operational KPIs, but these particular items are the essential factors that C-level management can closely track and use to manage change.
When mentioning to a chief executive officer that the MEL KPI is off by three percentage points, he might simply shrug and say, “So what?” Upper-level management cannot be expected to follow detailed KPIs closely.
However, the CEO can — and does — follow key analytics.
If, then, these are the more important KPIs, what more can be learned about where they come from and how they are processed and analyzed?
Where KPIs Come From
Airlines provide air service to passengers from Point A to Point B.
To understand the level of service provided to customers, we need to understand how the flights carrying those passengers perform. So the majority of the operational KPIs relate directly to the movement of aircraft. And the airline tracks aircraft movement through schedule times.
If the airline is using technology such as Sabre® AirCentre™ Movement Manager or Sabre® AirCentre™ Movement Control, approximately 90 percent of the data required to build these KPIs will come from these systems.
Alternatively, if the airline is not using one of these solutions, whatever flight-control tool the airline uses will most likely include these points of data.
The KPIs related to aircraft movement (departure performance, block performance, arrival performance and turnaround performance) follow a particular path for their creation and analysis.
In the data-processing flow, movement data comes to the flight-control system via ACARS, MVT messages or are input manually. The flight-control data stores the movement information and compares it to the schedule times. A delay is assigned for any flight that has a departure time later than the schedule departure time.
The flight-control tool also records cancellations as well as the reasons for the cancellations.
Next, in data storage, airlines pull data from the database contained in the flight-control tool. Some airlines do this once a day. Others pull (query) during the day to provide real-time performance updates.
The queried data typically goes into a data-warehouse environment. If formal data models are present in the data warehouse, the information coming from flight control can link to other tables containing relevant data such as flight plan, fuel consumption, weight-and-balance information, check-in information and passenger information.
Some airlines create a “table view,” in which all information is consolidated in one table (by flight or row). Depending on the sophistication of the airline’s system, calculations can occur while points of data are being published in the tables.
For example, with the time stamps for each flight (out, off, on, in), factors including departure delay, arrival delay and block difference can be quickly calculated. In performing these steps, counters should also be calculated — but not to the extent of generating unnecessary calculations against the database.
In data distribution and analysis, once the tables containing the operational data has been audited and closed (achieved through hourly, daily and monthly processes), the data can be used to report performance — including real-time, daily and other analyses.
To access the operational stored data, different business-intelligence tools or queries can be implemented. Sabre® AirCentre™ Operations Intelligence helps airlines that use flight-control applications from Sabre Airline Solutions® assemble these analyses, but other airlines may use either their own SQL or off-the-shelf applications such as Brio, Microsoft Access or SAP Business Object.
Even though these steps are generally quite straightforward and well known, they require strong business processes, clear definitions and solid audit processes.
Many airlines, for example, might take hours to input an “out” time. Other airlines may report based in Zulu. Therefore, when the performance-data information is published, these airlines are looking at a different snapshot depending on where that particular airline is located or in what time zone the majority of the airline’s flights are concentrated.
An issue that is essential to assure the relevance of KPI analysis is that the transactions registered in an airline’s systems genuinely reflect reality.
Again, strong business rules must be in place to register irregular operations such as diversions, ground returns, flight returns and cancellations. The quality of the KPI will always be directly related to the proper recording of these transactions.
Several examples of philosophical questions many airlines face might include:
- Is a diversion counted as a cancellation?
- What “out” time should be used in the case of a ground return — the first “out” time, which may very likely show no delay, or the second “out” time?
- If an entity owns two airlines and cancels one flight with Airline A but operates that flight with Airline B, is this a cancellation or not?
Airlines need to understand the limitations of particular data sources to understand the reliability of the KPIs.
A typical example is an airline using different methods to register out, off, on, in (OOOI) times. Some airlines have ACARS-equipped planes. Others have a mix of ACARS and non-ACARS.
And there may be an airline operating two fleets — one fleet with ACARS in which the “out” time definition is different from that of the airline’s other fleet (that is, the “out” definition for one fleet is brake release, while the “out” definition for the other fleet is last door close).
To keep the data as clean as possible, an airline needs a team of very specialized people who fully understand and assure that the data remains clean.
North American carriers generally employ dedicated statistical teams that work around the clock to validate and check for data accuracy. Other countries delegate this function to airport staff — with the airlines issuing clear guidance and policies for inputting and recording data.
With an understanding of how KPIs are created and where they are stored, airlines can establish operational goals based on the defined KPIs.
Block Hours Strategy
In addition to helping airlines establish and measure KPIs for a healthy operation, Manolo Centeno, a Sabre Airline Solutions consulting expert, has vast experience in many areas across an airline. Here, he discusses how an optimal block hours strategy is essential to an airline's operation, leading to increased profitability and improved on-time performance.
Establishing Operational Goals
KPIs must be mathematically driven and separated into groups by stations, divisions or types of processes (such as crew delays and other distinct descriptions). There should also be discernible relationships between individual KPIs to enable the clear documentation of dependencies.
To arrive on time, for example, a flight must depart on time. It must also have the necessary time interval to get from Point A to Point B. If an airline wants to achieve a new level of on-time performance on arrival, it is easy to calculate how much the departure performance must improve, assuming a given goal for block time has been set. This new expected level of departure performance becomes a goal for all departments involved in securing on-time departure.
In addition to the mathematical considerations, the following principles can be applied when developing operational performance metrics:
- The goals should be progressive and achievable. Senior leadership should balance aggressiveness against exactly how reasonably achievable these goals are. If the goals are unrealistic, people will not be motivated to believe in the goals, and change will not occur.
- The goals should be easy to split by division or outstations. Departure performance and completion factor are the best KPIs to apply in assigning goals at the station or division levels.
- Close attention must be paid to the geographic areas in which the most meaningful operational or business improvements are critical. If an airline, for example, has a strong profitability or revenue share in a specific region (such as Latin America or trans-Atlantic routes), the region should have its own set of goals.
- The goals should be rooted in a robust schedule. If the goals are to provide truly effective guidance to the business unit, the schedule must be robust. Once again, setting goals in an unrealistic and/or non-robust network will simply make the goals appear unrealistic to line employees. To make a network robust, the airline’s planning and operational planning must provide the right amount of block and ground time to operate the airline.
Measuring Network Health
KPIs should indicate to C-level executives whether the network is robust — meaning it is supporting, or at least can support, the flight schedule.
Because other resources are planned based on the flight schedule, if the network is not robust — or operations are not delivering — another set of problems is almost certain to develop, affecting operations even further.
KPIs help managers identify these trends so they can start the processes to correct the observed behavior.
To be effective, it is critical to conduct an end-to-end analysis of all KPIs. If only one KPI is analyzed, it cannot possibly — by itself — provide an end-to-end view on operational performance.
As demonstrated in Figure 2, the departure performance at 15 minutes is very good. However, the arrival performance at 14 minutes is 20 points below target. This is a clear indication that the network is losing time somewhere.
Maintain The Operation Without Further Delay Propagation
While the departure performance at 15 minutes is good, the arrival performance at 14 minutes missed the target by 20 points. Clearly, the network is losing time. The block performance, which is low, is likely the problem. In this instance, the airline has cancelled 5 percent of all its flights to avoid additional delays across its system.
A good point to examine closely is block performance, which is reported as low. Therefore, the only way to maintain the operation without further delay propagation is to cancel flights. In this example, the carrier has thereby canceled 5 percent of its flights. The high cancellation rate also could indicate that other problems exist such as lack of crew or problems with maintenance.
To return to a previous analogy, this is very similar to what the doctor sees when he or she analyzes your health results.
Acting as the “doctor,” an airline’s senior management can very quickly analyze and interpret the indicators of the true health and robustness of the operation.
Then the managers can start driving change and setting highly realistic goals to improve the operation. In turn, this translates into more satisfied customers and greater competitiveness for the airline.