Referenced in our Newsletter Volume 3, Issue 6 - June 2004
Analyzing Airline Profitability
This month's link chart is a bit non-traditional because the model does not contain any associations (linkages).
The data analyzed comes from an accounting system used by a leading airline to review their operating costs and the
profits/losses associated with their airline routes. Rather than focus on the individual flight costs that occur daily,
the type of data represented in this model is "summary" level because it rolls-up all the costs associated with the routes
over the most recent financial period. As can be seen in the sample record below, different categories are presented
(all amounts are represented in 1000s) including passive revenue, total depreciation, and fixed costs.
The entire model consists of a single "transactional" object type that supports all of the data provided from the accounting system.
The object created for the previous record (shown at right) has the flight number, origination location, destination location,
and equipment type shown for its label.
Upon loading all of the routes for this airline, the following clusters where shown, based on the aircraft equipment type,
in order to help determine the distribution of their operations. Due to the larger number of international (inter-continental)
flights supported by this airline, the diagram shows us they maintain a fleet of larger aircrafts (e.g., 767, 777, Airbus).
Next, we looked at how each of the values contained in the accounting data could be used to help predict the profit/loss
for the route. At this point a Parallel Coordinates placement was applied using PROFIT-LOSS, TOTAL-COST, and TOTAL-REV
as the initial dimensions. In the following diagram the objects representing each route are displayed at the far left
and the selected variables are shown normalized from low-to-high (bottom to top).
The lines shown between the variables (columns) depict the co-occurrence of values within one of the objects and if more
than one object has the same set of values, the linkages will be proportionately thicker. As can be clearly seen in this
diagram, a higher PROFIT-LOSS value often indicates a lower TOTAL-COST for operating that route which in turn equates to
a lower TOTAL-REV. Keep in mind, many of the values normalized along each column run from negative to positive numbers.
Based on the emerging pattern, the EQUIPMENT-TYPE was added to the diagram to determine if there is any type of dependency
on the size of the aircraft. The attribute was placed between PROFIT-LOSS and the TOTAL-COST and the diagram (shown below)
immediately showed that there were biases based on the equipment. The highlights (blue links) in the diagram show a smaller
aircraft frame that tends to show higher profitability and lower overall operating costs.
The next diagram shows highlights for a larger aircraft frame that represents a much larger cost infrastructure. Generally,
we saw that the larger the aircraft, the more it accounted for the total overall operating costs. Each equipment type was
reviewed to see where it fell in the overall corporate cost structure.
The final diagram shows the EQUIPMENT-TYPE by TOTAL-DEPRECIATION. Certain aircraft for this airline represent a much lower
epreciation cost, most likely due to the newness of their fleet. This information can help factor into the overall analysis
of how the airline allocates it fleet and routes to help minimize operating costs and maximize profit.
|
