Route and network profitability: how to measure it June 2000
This is the second in a series of articles on measuring an airline’s route and network profitability. The first, published in the February 2000 issue, focused on some fundamental development principles for management to adopt when setting out to build and/or improve this vital MIS tool. Assuming then — and it’s a big assumption — that the MIS development process is managed well, there remains a daunting list of technical, process and analytical challenges.
On the technical front, the strength of the internal IT group and the selection of experienced and successful systems and software suppliers — and contractors — will be a critical factor.
Key questions need to be resolved efficiently and cost–effectively, including how to: process/warehouse such huge volumes of data; integrate with existing architecture(s); meet data and information availability and timeliness goals, and deliver flexible online/ ad–hoc analysis.
The process challenges — alongside those outlined in the previous article — encapsulate difficulties such as:
- Ensuring accurate data capture and quality control at source — how many MIS tools suffer from the user criticism of "rubbish in, rubbish out"?
- Recognising the limitations of existing data capture processes, assessing these against the ideal MIS requirements and determining what and where to compromise; and
- The use and control of the MIS information by management.
The analytical challenge focuses on the robustness of the mathematical algorithms and allocation mechanisms used in building the route and network profitability model. This applies to costs and revenues which have to be combined to report profitability — along a variety of dimensions at different "reporting levels".
This article is structured around these four elements: dimensions, reporting levels, costs and revenues. Most airlines want visibility of profitability on a common set of dimensions: by flight number, by route, by city pair, by O&D, by hub, by fleet type, by geography/region and for the whole network. The specifics of the dimensions (e.g. which geography basis) are often driven by organisation structure.
The trick is to retain the flexibility in the model to absorb the inevitable changes in organisation (as new management arrive and consultants pontificate). Many airlines are going through the strains of migrating from a route–oriented approach to one based on O&Ds.
Since this MIS will inevitably be a, if not the, major repository of cost data, then managers may also wish to build in other interesting capabilities as well, to support decision- making. For example:
- Relative profitability of class of service;
- Cost effectiveness of different distribution channels; and
- Short haul versus long haul.
One final subtle dimension to be considered in designing the analytical framework is one of timeframe. The MIS will be used to assess historical route performance and support tactical decision–making. In this context and timeframe, the design can treat whole groups of costs as "fixed", for example, sales and marketing, flight operations overhead.
However, in a longer–term planning context, many of these costs become more "variable". So, in the case of the two examples given:
- Sales and marketing: significant long–term changes in network size or distribution may require staff costs in reservations, yield management and pricing to be modelled on a variable basis;
- Flight operations overhead: significant fleet rationalisation or reduction may require costs in operations control, central training and crew scheduling to be considered as variable.
Such timeframe considerations will impact the flow of data capture and management. The design and selection of levels to be reported should be driven by one simple question: can management take decisions and action based on the information presented?
We have seen airlines with the number of reporting levels varying from two to twelve. In the first instance, there was perhaps too little transparency; in the second, most of the levels presented were meaningless in terms of providing "actionable" information.
Analyses will be required at what may usefully be called activity levels:
- Passenger activity: to help understand the true incremental cost of carrying a passenger (e.g. meal, handling, transfer charges, in–flight services, and — for the really retentive — fuel burn);
- Flight activity: to help assess the profit earned by operating the flight, including the incremental costs incurred (fuel, landing fees, crew allowances etc.);
- Fleet activity: to understand the economics of operating the fleet, by including the incremental fleet costs (aircraft ownership, crew salaries and training etc.);
- Network activity: to identify network profitability by including items such as sales costs; and
- Airline activity: to measure profitability including all overhead costs.
Five levels are probably too many for primary, senior management reports but will be necessary for the real analysts who work daily with the system (e.g., network planners). This subject will be discussed in more detail in a future article. At the highest level of detail, route profitability reports require costs to be allocated to flight numbers. So the first thing to understand is the nature of existing data capture and how to make best use of it.
The purist approach, designed to deliver ultimate accuracy, would capture "actual costs by flight" e.g. actual catering charge incurred for the number and mix of passengers boarded. Such "perfect" information is rarely readily available and would be unnecessarily difficult and expensive to establish, so most airlines adopt an approach that uses either:
- Cost rates: a derived rate where the rate times a production driver gives the actual; or
- Fixed amount allocation: monthly total cost allocated using a valid cost driver.
Passenger and flight related costs (as described above) are usually generated and allocated using cost rates. Fleet, network and airline–related costs almost always require the allocation of fixed amounts (using management accounting values).
That is, the MIS design and usefulness will inevitably rely heavily on accurate capture of actual production values, particularly number of departures/legs and number of passengers. And that is not necessarily as easy as it seems (aircraft switches, cancelled flights, combined flights).
The allocation methods and calculation mechanisms themselves can become both complex and controversial. Solutions exist but have to be customised by airline because data capture capability varies, business priorities differ and users are prepared to make different compromises.
However, particular pitfalls and complexities that will be faced include the following.
- Capture and modelling of different passenger class costs;
- Dedicated check–in desks and associated impact on check–in staffing levels; and
- Customised handling facilities (e.g., meals, priority baggage or lounges).
- Differentation by time of day for landing fees, to reflect cost of operations at peak periods;
- Allocation of crew hotel and transport costs, to reflect the relative impact of operating and rostering a specific flight number with its directly associated hotel requirements;
- Handling/estimation of ATC charges (invoices from some countries are often received many months after flights are operated);
- Allocation of ground handling charges to departing and/or arriving flight numbers; and
- Allocation of higher line maintenance charges for flight patterns using longer ground times.
- Allocation of crew salary and training costs, to reflect the relative impact of operating a specific flight number with its associated crew rostering requirements;
- Whether to use actuals or unit costs for the allocation of aircraft ownership costs;
- Whether to use actuals or unit costs for "heavy" aircraft maintenance costs, given the seasonality of the activity;
- Ensuring transparency of utilisation effects during the year;
- Reflecting commercially–driven schedule decisions on aircraft utilisation (e.g., stay on ground down–line to ensure departure time at best time for market);
- Cost allocations for "tag" or utilisation–driven flights; and
- Allocation of crew salary and training costs where crews are dual–rated (as with the A320 and A330).
- Allocation of sales costs to specific routes or the entire network (e.g., the New York sales office not only sells seats on the New York to "home base" route, but also routes across network); and
- Should station overhead costs be distributed on a network basis or to specific routes?
- Should "overhead" type costs be allocated to specific routes at all?
- If so, what allocation driver should be used to avoid bias: block hours? RPKs? passengers? legs? directly attributable cost distribution? Daunted by the cost allocation problem? Amazingly, many airline managers consider this issue to be "relatively" simple — compared to the allocation of revenues. Certainly, revenue allocation tends to give rise to much more animated debate, especially in the US environment where hub flows and connections are key to revenue generation.
The situation is exacerbated by the complexity of revenue accounting and revenue information (interline, code–share, block space, unearned revenue, front and backend commissions).
Transparent and robust reporting of point to- point and connecting revenue mix is vital to understanding a route’s total network contribution, and will be discussed further in this series.
The complexity does not stop here. The profitability measurement process has to incorporate cargo as well. And what about O&D profitability as airlines move to O&Dbased management?