Aviation Network Planning – or: how the Crystal Ball works

This is an updated rewrite (not a copy) of my 2011 blog article.

When I actually wrote that one, there was news about Ryanair removing flights from some airports. As Michael O’Leary (the Ryanair boss) had already announced that Ryanair would ground aircraft in winter (previous year they grounded 80), that seemed to be a rather strategic move and O’Leary’s argument it would be because of high airport charges sounded much like a smoke-ball, decepting the media. For the airport on the other hand…? It gave me good reason to try to give an idea about the Crystal Ball used in airline network planning. And explain, why I say we “polish the Crystal Ball” at CheckIn.com, but leave it to the experts to look into it and interpret what they see ツ

Network planning always starts with an idea. Who brings the idea up is irrelevant, except it may be a good idea to do a bit additional research, if your boss does.

Having the idea, question is: Is it feasible? So as a route manager, you’ll talk to the airport, you check MIDT-statistics about flown passengers on the route or routes from neighboring (=competitive) airports, you check “facts”, such as official statistics on commercial relations between the two regions, attractiveness to potential tourists or “VFR”-traffic: Visiting Friends and Relatives. Statistical data…

The MIDT-data is by nature incomplete. Often it does not include direct sales with the airline, Internet sales, etc.. Even if it does, it is based on historic data that relies on facts that are meanwhile outdated. Have the commercial relations strengthened? Did they suffer from bad connection and have faded away? How much were, are and will the travelers be willing to pay for the ticket? If the flight failed, why … hard facts preferred, not guesswork or interpretations by the airports involved.

tlscatchmentsampleThere is something called “catchment area”, based usually on isochrones, frequently a wild guesswork by the airport itself, more wishful thinking than reality and without any indication on how the numbers where compiled.

In this example: 60/90 min then 300km? As the bird flies? The airport not even centered?

While working with Yulia on the basics of the CheckIn.com-isochrones, I found in all cases we checked discrepancy larger than 10% different from the airport’s own figures in the most simple calculations of passengers living in a certain distance to the airport. The majority was far off any reality. And in all but one case (Munich), the published “catchment area” did not cover at all that there are other airports competing with the traveler!

minden

An example: If you live in Minden, Germany you are about equally distanced to four airports: Hannover, Paderborn, Münster/Osnabrück, Dortmund and Bremen. For the CheckIn.com-isochrones, we take into account the size of the airport – number of passengers, as the more passengers, the more likely that people choose the larger airport. The distance – the closer it gets by driving time, the better. Other factors we consider relevant or only partially available – like flights to the destination and the weekly frequency and seats (the higher the more “competitive) will be added next. But quite some more! From that, the system now calculates “fully automatic” the catchment area. And the results are devastating! Even on a “global” airport scale.

The smallest discrepancy of the catchment given by the airport of how many potential travelers would use the airport to the catchment we calculated was 50%. The high ones +90%. That is just basic statistics…

Catchment Area Calculation Example

On top of it, you have the questions we did and do not yet take into consideration working on the Isochrones. Like VFR, like commercial relations, like tourism interest, like … And you better know, where you got the information from, how likely it is to be overly optimistic or (sometimes, rarely) pessimistic…

Having that information, you start fine-tuning. You likely have several possibilities, depending on the passenger potential. Take a bigger airplane with more (cheaper) seats and lower frequency. Or take a smaller one with higher frequency, what the business traveler’s will like, but the seat being more expensive. OECD gave some figures for a rough calculation I found quite reasonable for a first idea: An Airbus A320 or Boeing 737 with 150-180 seats costs about 8c (Eurocent) per seat-kilometer (not mile). Use this form to multiply that with the distance for yourself. That is the “net cost”, so add the (increasing) taxes and surcharges onto that cost to come up with an idea about the ticket price you need as an average for flying break-even…

A professional network planner takes into account much more accuracy, such as real cost of operations, variables depending on aircraft and total hours of weekly/monthly operation. But for a start…

first-flightNow comes the next tricky step: How many seats can you likely sell from A to B at this average price, and how many might connect to C but use your flight from A to B? Do you have interline or codeshare agreements at the destination? What air fare do you get? Do you operate the long-haul-sector as well? The cost there likely is cheaper as the aircraft is bigger. But you got to have an idea on connecting services – a science of its own…

So now, at the end of that exercise, you can calculate your risk – how many passengers are you sure about using the offered flights? If you are unsure, your risk is 100%. If you have 50% guaranteed load, you risk less. You also get support from the airports usually, either financial, or “marketing support”. Compared to your operational risk, these are “peanuts”. And in the beginning, you run high risk, so it is wise to not calculate that on your predictive models – if you can’t succeed without, you likely loose.

But then comes the key-point, most airports today fail to have an answer for. Having finally calculated the risk, who takes it? In most cases the airline, virtually alone! Under such circumstances, why should an airline risk a new route with questionable return of investment? In charter flights, the tour operators (sometimes) take the risk, also mostly on their own. Lately, unused seats on some business charter flights, operated by large corporations are being made available to the public, either via Internet or GDS (travel industry flight booking systems).

And if that works well, from such a “business charter” a “scheduled service” with a higher frequency can evolve. But in most cases, the airline is left alone.
Any question, why they limit their risk, by pulling the plug when the losses pile up beyond their worse case scenarios?

airline-dot-goneWith eroding net profits, eaten by kerosene, excessive flight crew salaries, taxes and fees, the ability to build a new route in two to three years is very limited. If the route does not take off immediately with a reasonable load factor, the losses pile up so fast, the airline risks extinction if it does not keep a tight control on it.

That’s a short summary into the “science” of airline route planning. If you are an airport who seeks a new route, better do your part of the homework. Can you convince your local “catchment area” population to support the flight, even if the schedule and fares in the beginning may feel higher than from one of the competitive big airports in reach? Will your travel agents sell your flights with fervor – I’ve seen travel agencies at airports offering flights from Frankfurt, Munich, Düsseldorf, but not from the regional airport they were living at! Do you have a grip on the local VFR market? And most important: Can you get commitments from your corporates, business associations (chambers of commerce) or politicians about guarantees on how many seats they could fill? Not just good intentions, but hard figures and financial commitments! If you are a touristic region: Can your tourism managers qualify, how many people they will have flying on the route? It is fantastic to see, how ingenuous most of the tourist boards are about such basic facts: “Fly and we’ll see”. Flown for a decade, not seen…

Okay. As I’ve put it in the past (as a German saying goes): I take a big long stick and grope in the dark. Aside of the statistics which we now provide at your fingertips it requires expertise, experience and good guesswork to do something with all that information you get. Good luck is part of the business.

To summarize planning of new routes:

  • The quality of figures available, as well as my trust in them.
  • The cost of operations
  • From those: A more or less realistic load factor expectation.
  • Who takes the risk?

The first two factors are subject to an educated guess…

And what about those airporst having lost their routes? They likely have to do above homework now, there obviously was a discrepancy between expectation and reality. And Ryanair? I honestly believe Ryanair simply cuts aircraft for winter – they may want to talk to others if the routes justify a year-round traffic! That may convince Ryanair to consider the next winter. Or do the above homework as well to match expectations and wishful dreaming with the harshness of reality.

Food for Thought
comments welcome

And please follow this blog! ツ