The benefits of hub-and-spoke networks November 1998

Airline networks and schedules are increasingly being seen as key marketing tools in their own right, and one of the most important changes to airline operations in recent years has been the shift towards hub and spoke networks. Here, in the first of two articles on hubbing, Dr Nigel Dennis, senior research fellow at the University of Westminster’s Transport Studies Group, examines why hubbing is so important and why it lies at the centre of any attempt to maximise the potential of an airline network.

Increase in market coverage

The most immediate benefit of hub and spoke networks is to increase greatly the number of city pair markets that an airline can serve for a given volume of output. Consolidating many different traffic flows together through a hub can thus offer a very efficient means of relating supply to demand.

Minimising the transfer time

If the passenger is prepared to wait an indefinite time at the hub, connections can be achieved between all services operating to and from it. In reality, long delays at the transfer airport are unattractive especially where the actual flying time is short. If alternative routes are available, a considerable drain of traffic may be experienced (for every 30 minutes spent on the ground, the passenger could fly another 400km).

An essential element of any serious attempt to maximise the scope of an airport as a hub is to concentrate activity into a limited number of peaks or waves during the day. These should see a large number of inbound flights arriving in a short space of time, then departing again as soon as a sufficient interval in which to redistribute passengers and their luggage has elapsed.

Although the volume of flights at a busy airport such as Heathrow ensures that many connection possibilities will exist by chance, it is only through operating waves of flights that a consistent connecting timetable can be provided, with services in both directions in each city-pair market and a transfer time close to the optimal. But high frequencies are not a prerequisite for hubbing. Indeed at many US hubs only about three flights per day are operated on most routes.

Costs will increase as a consequence of creating these artificial peaks of activity but this can be offset through the economies of consolidating traffic onto larger aircraft or operating at higher load factors. The marketing benefits are potentially much greater.

Elimination of interlining

Commercial agreements between airlines have been a major component of regulation in air passenger transport. Multilateral interline procedures were recognised as being in both the operators’ and the public’s interest. For the airlines it was seen as essential to attract business that they could not otherwise serve.

The demise of these traditional arrangements has been most marked since deregulation in the US. Whereas half the passengers changing aircraft in the US in 1977 also changed airlines, this figure has fallen to less than 10% today. In Europe too, on-line or code-share connections are increasingly dominating the market. Whereas the proportion of transfers at Heathrow that were BA–BA was only 27% in 1984, this had risen to 43% by 1991 and is nearer 60% today. BA–BA pairings, however, account for only about 16% of the possible linkages at Heathrow. This means BA–BA transfers sell on average six times better than those involving any other pairing of airlines. If one further removed code-share connections such as those between British Midland and various carriers, the remaining interline transfers such as AF–BA or SK–AA are clearly little used.

The reasons behind this shift to hubbing are as follows: without restrictions on route entry, airlines have been able to enter markets previously closed to them. By routing these services through a common hub, on-line travel can be provided. Furthermore, waving of flight schedules ensures that the probability of the first outgoing service to any particular destination being by the same airline as the delivering flight is disproportionately high. Consequently, it no longer becomes necessary for airlines to offer interline-able fares in many important markets, since it is possible for them to supply an optimal service of their own.

Subsequent expansion of point-to-point services in the US by airlines such as Southwest has been largely counter-balanced by the disappearance of Eastern in the N.E.-Florida market and Air Cal and PSA in California. Other new entrants such as America West, Midwest Express, Reno Air and even Air Tran (the renamed Valujet) operate essentially hubbed networks. In Europe, Virgin Express is a hub-and-spoke operation. The Majors have tended to divert the resources from merged airlines to strengthen their hubs. It should, however, also be noted that hubbing has not led to a huge switch from direct to indirect travel. Although some non-hub cities have lost certain non-stop links, many new non-stop flights have become available from the hub cities themselves.

An extension of the on-line connection concept involves bilateral interlining with complementary carriers. This has grown considerably in recent years, assisted by devices such as code-sharing. It is thus increasingly individual airlines, or groups of airlines, that form a hub at a particular location. The traditional concept of a hub simply as a large airport is no longer very valid.

Maximising the number of marketable connections: directional hubs

It is apparent that not all possible connections through a hub will be of value. Where a significant back-track is involved, passengers are likely to be deterred by the increased flying time while airlines may be unable to offer a viable fare by the circuitous route.

Connections within a hub wave will be universally good while those between waves will be relatively poor. Although the greatest number of linkages would be achieved by concentrating all activity into one or two huge waves each day, this is usually impractical. The aim therefore is to reduce the number of flights in each wave while ensuring as far as possible that it is the least marketable linkages which are lost. This can be achieved by seeking sub-groups within the set of routes operated from the airport between which there is a major demand for connecting travel but within which there is not. Whereas with traditional scheduling methods, aircraft return back on the same route from which they originated, they should now proceed on through the hub to a location from the contrasting set. This ensures that all the immediate connections will be marketable, which cannot be achieved with a random timetable and maximises the efficiency of the hub for any given level of resources.

The most straightforward separation that may be adopted is to introduce a geographical orientation such as East–West so that flights from one region operate through the hub to points broadly in the opposite direction beyond it — so-called hourglass hubs. It is demonstrated in the classic East–West hubs of the US such as Chicago, St Louis and Dallas. The schedules facilitate journeys such as Boston–San Diego or Miami–Seattle but not Boston–Miami or San Diego–Seattle. In Europe, Copenhagen (Scandinavia–Europe) and Vienna (East–West) follow this pattern, albeit on a smaller scale.

If this arrangement is not appropriate, the principal alternative is a differentiation by length of route. This features short sectors operated between the hub and nearby cities in order to generate feed for the longer distance trunk routes. As one stage of the journey is much longer than the other, the hub can become multi-directional for connections between these groups as back-tracks and dog-legs will not be of significance. These can be described as hinterland hubs because the central airport serves as a distribution point for air travel to and from its surrounding catchment area.

There are several examples of “niche” hubs in the US following this pattern — Midwest Express at Milwaukee and the former USAir hub at Dayton — while in Europe, a number of airports such as Amsterdam and Zürich are primarily aimed at being interfaces between short-haul and long-haul flights.