Development

Isochrones are not Catchment Areas

Today, I had a lengthy discussion about why we give away isochrones maps for free. I used three images to visualize quickly where we came from and why we consider isochrones maps useless and such something we give away.

The Classic Isochrones Map

Classic Isochrones Map
Classic Isochrones Map

The classic Isochrones map (here Hamburg) shows how far you get driving a certain time. In case of Hamburg we look at 60, 90 and 120 minutes. We differentiate the “reach” by the number of annual passengers the airport has.This has become a standard to display, though many airports show rough estimates of “like the bird flies”, they use rough numbers. Where we have sound drivetime calculations (mostly based on OpenStreetMap, with spot checks on more than 20 other tools and quite some corrections sometimes) for any airport in Europe.

These isochrones analyses were offered at prices of €3500 or more, with waiting times of three to six weeks. Only big airports worked with more or less sound statistics, there was a lot of guessworking. We had frequent differences in excess of 50% from our calculations, that’s beyond “drivetime differences”, Munich we matched better than 95%. The smaller the airport, the more often we think their numbers were “guestimates” and/or very outdated numbers.

Adding the Competitive Airports

Isochrone Map with Competitive Airports
Isochrone Map with Competitive Airports

Once receiving isochrones with the population from an airport, as a first step we added airports to it, to understand the competitive situation and adjust the gross population somewhat. That sure had to be mostly guesswork and the numbers where anything but “scientific” or “sound”.

But do you truly believe that someone living directly in Hanover will travel all the way to Hamburg to fly? The passengers’ choice – and the passenger does not live in Hamburg, but up to a few hours away and may have more than one airport to chose from…

That was the trigger for the development of CheckIn.com.

The Catchment Area

Catchment Area Analysis
Catchment Area Analysis under consideration of the impact of the surrounding airports

Based on that approach, we sought existing solutions but did not find any. If solutions were offered, we were told it would take four to six weeks for such an analysis, not cross border and insane amounts of money. Such we decided to develop a database to calculate “real” catchment areas.

Having ourselves backgrounds with small and midsized airlines and airports, with the prices being asked for simple isochrones, we also decided intentionally to make our offers available at reasonable rates to promote better usage of such data and less reliance on guesswork. Isochrones we give away for free, a catchment area analysis is offered at € 1.200. We currently work to add a route level analysis, which we will offer for € 1800, until then airports or airlines can already order the two airports’ catchment areas and once we switch the route level, you get that on top.

Isochrones are not Catchment Areas

With competitive airports around, depending on their size (usually also reflecting the number of flights and destinations) and the distance to where the traveler lives, the isochrones map looks appealing, the numbers are high, but they are far from any commercial reality. That’s why we give them away for free. Trash them. The underlying drivetimes are valuable for more indepth calculations. In themselves, they are useless!

P.S.: Another Example: Denmark

 Working on custom colored maps for TheRouteShop (for free), there were four Danish airports there, AAL, BLL, CPH and SGD. Looking at the catchment, they so nicely overlap. So I had a look at how they compute with the function we developed for our customers using “airport systems”, where we i.e. show the combined catchment area using a common color. So I scripted the four airports to show different colors. The analysis takes all airports in that region into account, nicely visible at Aarhus at the center “light pink” area.

This is nothing we can provide commonly, this is example was straining the unusually extensive resources of our mapping server to the limits. But yes, if you are willing to purchase the different analysis, we may be able to do some additional compilations (at modest charges). On request and not guaranteed, I happen to crash the test server trying out a similar map BSL, GVA, ZRH, BRN…

1

Leave a Reply

Your email address will not be published. Required fields are marked *