The end of growth for low-cost airline networks? : route churn as measure for route density problems in airline networks

(1)

UNIVERSITEIT VAN AMSTERDAM

The end of growth for

low-cost airline networks?

Route churn as measure for route density problems

in airline networks

Author: Nina Moers

Student number: 10664920 Hand-in date: 14-08 2014

Thesis supervisor: Prof. Drs. J.G. de Wit Second corrector: Drs. E. Dirksen

Faculty of Economics and Business Studies Master thesis

(2)

1

Abstract

The European airline market shows increasing signs of saturation, leading to problems in the network development of European low-cost airlines. The objective of this thesis was to prove an increase in route churn over the years and to give a characterization of these churned routes. An increasing amount of churn in the network was said to be caused by problems with route density. In this thesis, a difference was made between three different types of low-cost models. It was expected that churn would be especially relevant for ultra low-cost carriers as their focus on secondary and tertiary airport types with low catchment areas creates a network with a high degree of density routes. This was illustrated by taking the case of ultra low-cost airline Ryanair as a prime example. A database including information on 7486 routes operated by Ryanair, EasyJet, Norwegian and Wizz Air between the seasons 2002/2003 to 2013/2014 was used. Regression analyses revealed a positive relation between churn and network size and a negative relation between route discontinuity and operation time. Qualitative analyses provided per airline an overview of the type of routes with the highest chance on churn. Support was found that the ultra low-cost model has the highest level of route churn. Indeed, their problems are so severe that even their network sizes declined in season 2013/2014.

The outcomes of this study give insight in the current and future network development of European low-cost airlines. Moreover, the increased churn rate functions as additional evidence that the European airline market shows increasing symptoms of saturation.

Key words: Low-cost carrier, Ryanair, route density, route churn, market saturation, network development

(3)

2

1.0 Introduction

The revolution of the low-cost carrier (LCC) began with Southwest in the US market in the 1970s, and has shown to have a big impact on the airline industry. Before the rise of the low-cost concept the market was dominated by full-service carriers (FSCs). These FSCs were offering an integrated product which included the whole flight experience for one price. On the contrary, LCCs completely unbundled the product offered by FSCs. They ask a price for a seat on their plane, and all additional aspects of a flight, e.g. meals, priority check-in, seat allocation, luggage etc., can be bought separately.

An insight into the statistics of members of the European Low Fares Airline Association (ELFAA) shows that Ryanair and EasyJet are the biggest low-cost operators in the European market with regards to passenger volume, followed by Norwegian, Vueling and Wizz Air. An overview of this and of their fleet- and network-size is shown in the following table.

Airline Home country Passengers (Jan-Dec ’13)

Number of aircraft (Dec ’13)

Number of routes (Dec ’13)

Ryanair United Kingdom 81.4 m 303 1600

EasyJet United Kingdom 61.4 m 217 689

Norwegian Norway 20.7 m 86 395

Vueling Spain 17.2 m 66 252

Wizz Air Hungary 13.5 m 45 303

Table 1: Overview 5 biggest ELFAA members Source: ELFAA (2013)

Ryanair and EasyJet have known exponential growth and good profit numbers since the beginning. Though both are considered as a low-cost airline, their business models know significant differences. Both airlines operate on a point-to-point route network which allows fast turnaround times and cost cutting on expensive hub-operations. They differ in that Ryanair is by many considered as the prime example of a ‘lowest-cost’ or ultra low-cost

(6)

5 carrier (ULCC) (Tretheway, 2004; European Cockpit Association, 2006), by offering routes to secondary and tertiary airports and having a ‘no-frills’ service. EasyJet has a so termed ‘hybrid’ airline business model (Daft & Albers, 2013), since it offers a low price, but provides additional service features – previously associated with FSCs – such as destinations to primary airports and seat selection.

Over the years the airline industry has become a highly competitive environment with a relatively low level of financial performance compared to other industries. A report from the International Air Transport Association (IATA)(2011) found that on average, LCCs are more profitable than FSCs. However, carrying out a low-cost strategy does not ensure these profits, as is proven by the many low-cost airlines that entered the market and dropped out again after they could not achieve profitable operations (Budd, Francis, Humphreys, & Ison, 2014). There does not seem to be a certain size, business model, or geographical area that explains why certain airlines were doing better than others.

Indeed, the airline sector is a challenging market, both for FSCs and LCCs. One of the current challenges identified by a study of de Wit and Zuidberg (2012) is the increasing problem LCCs have with route density caused by a saturating market. Growth becomes increasingly hard for LCCs as they have expanded their networks over the years so widely that, after all routes with high density were served, growth was only possible if they started to expand to less dense routes as well. Inevitably a point must come that most of the exploitable routes between city pairs have already been captured, and the upper-limit of the number of exploitable routes is reached. This observation made de Wit and Zuidberg (2012) point out the question ‘are there growth limits to the business model of the LCC?’, a question that becomes very relevant and prominent with regards to this saturating market.

(7)

6 Problems with route density will be illustrated in this thesis by taking the Ryanair case. The airline has had a strategy for many years of operating predominantly on secondary and tertiary airports, thereby trying to establish network growth by the continuous addition of low-density routes. In this search for additional routes, thinner routes are tested for a minimal usage to see if airplanes with a minimal occupation level can be used. However, the test of these low-density routes often results in a high amount of ceased routes and operating bases due to insufficient demand. An operating base is an airport where a carrier has fleet and crew permanently based and from where multiple routes are operated (CAPA, 2013a). A consistent number of aircraft and crew return to the base every day, thereby enabling the hiring of local cabin crew. In the summer of 2014 Ryanair stopped operating from a substantial number of bases and cancelled 222 existing routes (Anna Aero, 2014). In concrete, the airline cut the number of routes operated for the first time in its history, as the number of ceased routes surpassed the number of newly added routes this year. Therewith, Europe’s biggest low-cost operator is showing a declining route network in the current flight season (Anna Aero, 2014). Understanding the magnitude and growth of these route variations might give insight into the route density problems for LCCs.

The result of these route variations is that Ryanair’s network becomes more instable over the years. As response to this network instability Ryanair is introducing a new strategy direction. The airline announced recently to start operations at primary airports in e.g. Brussels, Rome, Milan and Madrid (Ryanair, 2014a), which is in contradiction with lowest-cost principles. Thus, the announcement of primary airport usage seems to demand a shift towards a less ultra low-cost business model, leaving its single focus on secondary and tertiary airports behind. The question is if this changing strategy will solve their problems with regards to a saturating market for lowest-cost carriers.

(8)

7 In this context, Ryanair’s move towards primary airports is explained as an escape from route density constraints. This thesis contributes to existing literature by providing insight in these route density problems by introducing a new concept from the marketing literature, churn, to measure the variation in routes started and ceased over a certain time period. A route is said to have churn when the route is started but ended again after a relatively short time period. Thus, if an airline has a high churn rate, this might explain a heavy search for exploitable point-to-point destinations and thus a problem with route density, or it is the indirect cause of strategy adaptations by shifting routes to primary airports.

In this study the concept ‘churn’ is used to measure the amount of route churn within the networks of 4 European low-cost carriers. It functions as an indicator of the increasingly difficult search for exploitable point-to-point routes by airlines. The results can be used as complementing evidence that market saturation is becoming an increasingly bigger problem for LCCs, and for Ryanair in particular, and that the business model of the lowest-cost operator eventually has limited growth possibilities. Furthermore, the objective of this study is to give a characterization of routes with churn, and to predict which routes have the highest chance on churn based on airport type, region operated in, airline strategy, and operation time of a route. In addition, this thesis discusses how airlines respond to route density problems by optimizing their business models and how their networks develop to lower the chances on route churn.

This study proceeds as follows; in the next section the literature on low-cost carriers will be presented. Further, the expected link between route density problems and market saturation is discussed, as well as how the introduction of the churn rate can give insight in the magnitude of these problems. In the third section, four selected airline models are described, followed by the research design in section four. Subsequently the analysis and results will be addressed, followed by a discussion and ideas for future research.

(9)

8

2.0 Literature review

2.1 Competition airline industry

Until the late 1970s the European airline business was a highly regulated industry where most airlines were publically owned and often enjoyed government subsidies. However, the liberalization of the sector that started in the late 1970s was the start sign for many new players to enter the market and compete for a share of the profits. By then the US market, which was mostly dominated by privately owned carriers, had already undergone extensive deregulations. It was in this period that the first successful low-cost strategy was applied by Southwest in the United States, a concept that came to Europe in 1995 with the introduction of that model by Ryanair. After the European market became liberalized, European airlines were allowed to fly and start operating bases within any other EU country and to fly between any country pair within the EU. This enabled high flexibility in the route development of point to point carriers such as LCCs.

Traditional legacy carriers were very much affected by the rise and popularity of the LCC concept. Most of them saw their profits and market share drop heavily over the past years. Indeed, the airline sector is a very tough market, with a relatively low level of financial performance compared to other markets (IATA, 2011). The sector is not only affected by an increasing price pressure caused by LCCs, but it also copes with high fuel prices, volatile foreign exchange rates and an uncertain demand, which makes the airline industry increasingly challenging over the years (Graham, 2013; Merkert & Gudmundsson, 2013). Airlines, and legacy carriers in specific, have to work harder to maintain their market share and profits.

(10)

9 2.2 Airline business models

According to Porter (1985), a firm can pursue three generic strategies to create an advantage over its competitors and to guard itself against the five competitive forces: either follow a differentiation strategy, pursue a low-cost strategy or implement a focus strategy. The first two strategies are discussed here, as they are most widely applied in the aviation industry. In the first strategy, differentiation, a product or service is offered that entails unique attributes that are valued by the target market and allow the firm to charge a premium price for the product or service. This strategy is often associated with traditional airlines or the FSC model. These airlines apply the so called hub-and-spoke model. They build their networks around one or a few big ‘hubs’ with a large capacity utilization, which are the link in a network of feeder routes. The structure of such a network enables the exploitation of economies of scope and density (Burghouwt & de Wit, 2005). The operational complexity of these hubs is very high, as it comes together with for example costly baggage handling, long turnaround times and high labour costs at the hub (de Wit & Zuidberg, 2012). The passengers targeted by FSCs are travellers who need high inter-flight flexibility and ground service linkages (Hunter, 2006). Therewith, the extensive long-haul intercontinental networks of FSCs require European flights to be connected as good as possible. Frequent flier programs, major city airport destinations and comfort on-board are examples of differentiating attributes that allow FSCs to charge a premium price.

The second strategy that a firm can pursue according to Porter (1985) is the implementation of a low-cost strategy. This alternative strategy requires a company to compete on the basis of costs, or to create competitive advantage by offering its products or services for a lower price than the competition. While the term low-cost carrier is frequently used, there is not a uniform description of its characteristics and the term is used to describe a variation of models which tend to have low-cost aspects (Button, 2012). Generally, LCCs are

(11)

10 characterized by offering a ‘no frills - low price’ product. They achieve significantly lower costs by providing a small amount of service and only single class cabin. Moreover, their operational bases are established with a homogenous fleet predominantly on secondary or tertiary airports where landing charges are much lower (Doganis, 2010). Most importantly, to eliminate hub costs and to keep their operational efficiency as high as possible, LCCs work with a point-to-point system where only direct flights are offered (Lenartowicz, Mason, & Foster, 2013). Moreover, direct and online booking and the absence of a network with feeder routes have enabled high cost cutting for those airlines (Hunter, 2006). These lower prices enabled them to lure passengers away from incumbent FSCs, but also made them generate new customer groups who previously could not afford traveling by air (ECA, 2006).

Since the rise of the low-cost concept, deviations in the fundamental principles have emerged where LCCs try to adapt their business models to offer a combination of both differentiation and low-cost, the so called ‘hybrid’ business model (Alamdari & Fagan, 2005). The ECA (2006) describes this phenomenon by distinguishing between ‘core’ low-cost and low-cost airlines, where the latter are compared with the hybrid business model. An assessment of the characteristics of this hybrid model is given by Klophaus, Conrady and Fichert (2012). They made a comparison between the business models of the 20 largest European LCCs and assessed to which extent LCCs follow the archetypical low-cost characteristics or blend their model with features of FSCs. Their study leads to a classification of four different low-cost business model types: Pure LCCs, hybrid carriers with dominating LCC characteristics, hybrid carriers with dominating FSC characteristics, and the full-service airlines. The classification is based on counting how many of 13 predefined low-cost criteria are fulfilled. These criteria are shown in the following table.

(12)

11

Table 2: Criteria for the LCC business model. Source: Klophaus, Conrady & Fichert, 2012.

The pure LCC model is operated by airlines that meet at least 11 criteria of a low-cost airline. Airlines that are called hybrid carriers with dominating LCC characteristics obtain a score between 8 and 10 low-cost criteria. When only between the 4 and 7 criteria are met an airline is defined as a hybrid carrier with dominating FSC characteristics. Finally, airlines that are categorized in the last type and have a score of less than 3 low-cost criteria may not even be classified as hybrid airlines, and based on their criteria do not fall into the low-cost airline group. This classification infers that within the LCC sector many airlines have diverging market strategies.

2.3 Prognoses airline industry

The market share of LCCs grew rapidly since they started operating in the European market in the 1990s. FSCs were forced to rethink and adapt their short-haul strategies in order to maintain (part of) their market shares. At the beginning of the century, forecasts for European low-cost market share for 2010 varied between the 25 to 33% (Campbell & Kingsley Jones, 2002). By 2010, however, European LCCs had reached a combined market share of 39% (IATA, 2010), thereby outperforming everyone’s expectations. For many years, prognoses for the sector had been in favour of LCCs, forecasting untrammelled growth partly at the expense of market share of FSCs. A report of York Aviation that was published in 2011 forecasted a

(13)

12 double-speed growth rate for European LCCs in the ten years after publication, as well as an intra-European passenger market share for LCCs of 53% in 2020 (York Aviation, 2011).

However, prognoses with regards to market shares are dependent on which airlines are considered to belong to the LCC segment. Much literature has focused on the nature and definition of the low-cost model, yet as low-cost models continuously develop and evolve no unambiguous assessment exists. For instance, the abovementioned classification of Klophaus et al. (2012) is based on an assessment on what level a LCC follows the archetypical low-cost model, or how much it deviates towards FSC features. The classification made by Francis, Humphreys, Ison and Aicken (2006), on the other hand, looks at the origin of a carrier and classifies airlines into five typologies based on the criteria whether it started from scratch or on the type of carrier it was before it entered the low-cost segment. Comparably, the list off LCCs that are included in CAPA (2014) databases is much longer than the LCCs identified in research by ELFAA (York Aviation, 2011), where ELFAA is much more strict in which airline to assign a LCC status or not. Therefore, accurate predictions with regards to European low-cost market shares remain somewhat disputable. In this regard, the question posed by Budd et al. (2014) about what type of airlines enter the low-cost market and which type of airlines have the highest probability of exiting again becomes relevant. Their research showed that in the period between 1992 and 2012 77% of entrants failed and exited the market again. Airlines that, among others, stuck to the original LCC model and had first mover advantage were found to have a higher chance to survive. LCCs that have a smaller scale and scope of operations due to small fleet size and less resources are typically the airlines that fail (Budd et al., 2014). Giving the increasingly mature market in Europe they predict that opportunities for new LCCs are limited and that growth for established LCCs will not be substantial. Indeed, latest predictions concerning LCC development are becoming more cautious. Klophaus et al. (2012) still describe an increasing market share for low-cost airlines, but see passenger

(14)

13 growth slowing down as European markets are becoming more mature. CAPA (2014a), the market analyst of the airline sector, even indicated that the low-cost market within Europe has been stagnating over the past 4 years (graph 1).

Graph 1: Total capacity share of LCCs within Europe Source: CAPA, 2014

Additional prove on this development is given in an article of Leigh (2014). He shows that the biggest LCC market, namely the UK-Europe market, has matured and the growth trend over the past 13 years has been downward, trending towards an organic growth rate of 2% (graph 2). In line with these developments, de Wit and Zuidberg (2012) raised the question if the European market for low-cost airlines is limited and finds itself in a saturating phase.

(15)

14

Graph 2: Growth trend LCCs in the UK-Europe market Source: Leigh, 2014

2.4 Route seasonality

One of the issues that helps to explain a slow-down in the growth of LCCs is seasonality. Seasonality has always been a very relevant topic in the airline sector. It occurs when a city pair copes with peak-demand during a specific time period in the year, while there is an off-peak demand during the rest of the year. The importance of seasonality is especially high in tourist markets (Garrigos-Simon, Narangajavana, & Gil-Pechuan, 2010), and thus particularly important for airlines focused on leisure travellers instead of business travellers. Causes of seasonality in tourism vary between natural causes, such as climate and weather in a specific time of the year, and institutional causes, such as public holidays and school holidays (Boffa & Succurro, 2012).

For airlines seasonality is an important market characteristic, as this creates an unbalance in aircraft usage and pricing. During off-peak season, prices are decreased and flight frequency and seat capacity are reduced during this period to respond to lower demand. As a consequence, the aircraft fleet is not optimally deployed during this period, unless replacement routes can be found. Finding replacement, however, might not be so easy for intra-European traffic, as the natural and institutional causes of seasonality are roughly the same within Europe. Generally, this causes more demand in summertime for intra-European

(16)

15 air traffic than in wintertime. Moreover, when airline networks expand and fleet sizes grow, the absolute rate of under-capacity during off-peak seasons grows with it, creating a higher amount of aircraft not deployed off-season. The OAG database used in this study shows that under-capacity during off-peak seasons has increased significantly in absolute terms for four LCCs (Ryanair, EasyJet, Norwegian and Wizz Air) in the time period between 2002 and 2013 (graph 3).

Graph 3: Frequency difference between summer- and winterseason for selected airlines Source: OAG database 2014

In this graph the difference in flight frequency between the winter and the summer season is shown per airline. The relation between airline size and seasonality is clearly demonstrated here, as the graph makes clear that the bigger an airline becomes, the bigger the difference in route frequency between the winter season and the summer season gets. Especially Ryanair has, due to its size, a very big gap between its flight frequency in summer and winter time. These differences lead to an unstable network with much variation between the two seasons. Besides this frequency gap, Ryanair shows - from 2011 on - a gap in seat capacity that more than doubles in the summer season compared with the winter season. Consequently, a smaller part of Ryanair’s fleet will be deployed during off-peak season due to substantial overcapacity of the fleet. Indeed, Ryanair grounds over 20% of its fleet every

-50000 0 50000 100000 150000 200000 20 02 20 03 20 04 20 05 20 06 20 07 20 08 20 09 20 10 20 11 20 12 20 13 Fr eque nc y ro ut es to ta l

Frequency difference between seasons

Ryanair EasyJet Norwegian Wizz Air

(17)

16 winter season since 2011 (CAPA, 2013b; The Telegraph, 2012; Ryanair, 2013a). Apparently, a part of Ryanair’s network only meets demand during the summer season, and there are no substitute routes available during the winter season. Remarkable is that Wizz Air shows a significant drop in frequency difference in the last year, in which the line even drops below zero. This indicates that the flight frequency was higher during the winter season for the airline, contrary to the tendency of having a dominant summer season. This drop points to a more balanced frequency between the seasons, leading to a more stable network for the airline.

Keeping airplanes grounded is a very costly and thus undesired operation. Seasonal grounding avoids variable operating costs such as fuel expenses and airport handling costs, but it does not reduce the company’s fixed costs like aircraft ownership, lease payments, rent of buildings, airport parking charges and employee costs. Furthermore, the company suffers from lost revenues that would have been made when an airplane could have been ‘normally’ deployed. Therefore, the question is raised what airlines have to do with their unused fleet during off-peak season, or how to avoid it. As graph 3 suggests, bigger low-cost airlines within Europe face heavier problems in absolute terms with seasonality than smaller airlines do. As such, one might wonder if the market in Europe for LCCs is limited, and if seasonality is one of the limiting factors of growth for LCCs.

Obviously Ryanair recognizes the consequences of the high level of route seasonality in its network and the necessity to decrease this. As a response to the seasonality problems the airline announced in its 2014 quarterly reports that it will start to focus on primary airports, as these airports promise to provide more stable year-round markets (Ryanair, 2014b &c).

(18)

17 2.5 Market saturation

In addition to the seasonality issue, more problems in the sector can be identified. The report of CAPA (2014) and the article of Leigh (2014) on the UK-Europe market revealed a market stagnation in which growth for LCCs is only limited. Their growth is increasingly challenged by route density problems (Graham, 2013). De Wit & Zuidberg (2012, p.17) define route density as “how many people you can book on an airplane at the same time, to the same destination, at a combination of fares that will ultimately cover the total costs of operating the flight”. The low-cost structure of LCCs enabled these airlines to offer flights to routes with a density that was too low for FSCs to operate at (Dennis, 2007), thereby extending the European route network. However, even for LCCs there exists a critical minimum density level on routes that is necessary to keep operations profitable, and as such not every route qualifies to be included in the network. De Wit and Zuidberg (2012) argue that a stagnation in growth of LCCs is caused by this route density problem. In the continuous effort of network expansion, finding new, exploitable routes becomes more difficult, as there are fewer of them left. Besides, seasonality increases these difficulties in finding new routes that could be operated profitably. Several signs express this market stagnation, one of them being a severe decrease in average flight frequency over the years reflecting an ever increasing share of low density routes in the network of ULCCs.

2.5.1 Frequency decrease

The minimum density level of a route is determined by aircraft type and flight frequency, as aircraft size determines the minimum amount of passengers needed to obtain a sufficient load factor and flight frequency indicates the minimum amount of weekly operations. De Wit and Zuidberg (2012) found that weekly flight frequencies of LCCs declined in the period between 2001 and 2010, while the number of routes served increased in the same timeframe. These results indicate that LCCs shift their attention to smaller markets with lower frequencies, as

(19)

18 larger, unserved city pairs seem to be hardly left over. Indeed, as all airlines considered in the research demonstrate the same patterns, the problems with route density support the conclusion that market saturation is taking place in the European market for business models that focus on secondary airports (ULCCs). Of the five airlines included in their research, Ryanair shows the most significant results. The airline shows a declining weekly route frequency from 16.4 in 2001 to 5.0 in 2010, while their number of routes grows from less than 200 in 2001 to more than 2000 in 2010 (de Wit & Zuidberg, 2012). These numbers mark the high amount of thin routes within Ryanair’s network.

This issue is reinforced on thin routes by the size of the aircraft operated, considering that airlines with a LCC formula often choose to operate with relatively big medium-haul aircraft like the Boeing 737 or the Airbus 320 series. This is because aircraft type is paired with a decreasing cost per seat for bigger aircraft. Ryanair for instance operates with Boeing 737-800s which are configured for 189 passengers. However, on thin routes is it almost impossible to fly daily with 189 seats, and it is questionable if a break-even load factor can be reached with a lower frequency. Such route density constraints might negatively affect Ryanair’s growth potential for the future, slowing down the tremendous growth they showed since the airline started in 1995.

2.5.2 Increased airport competition and density

Another troubling development and sign of market saturation is the increased level of competition among airports served by the LCC sector and the increasing airport density. The increasing competition affects market shares and bargaining power at airports (Graham, 2013). Not only competition from airlines operating at the same airports plays an important role, but also competition from neighbouring airports becomes a bigger issue. As the European route network is extended with more routes, more underused airports are served in the LCC networks and as a consequence the average distance between airports served in the

(20)

19 same network is decreasing. This increases the chance that catchment areas will overlap and increasingly cannibalise their own catchment areas by expanding their networks, as was for instance the case when Ryanair opened operational bases in Groningen and Maastricht in addition to Eindhoven, its main base in the Netherlands (de Wit & Zuidberg, 2012). Due to this overlap, common destinations will likely be cannibalised by each other. This phenomenon is another sign that low-cost airlines have difficulties to maintain substantial growth, and that their growth is only limited (Leigh, 2014).

In conclusion, in this increasingly heavier search for new city pairs, more routes start to fail. Thus, market saturation leads to an increasing discontinuity of routes. This creates high route volatility and network instability for the airline. Insight in the level of route discontinuity will also clarify the gravity of problems with market saturation. Therefore, the concept of route churn will be introduced to be able to measure the level of route discontinuity and draw conclusions on the stability of the network.

2.6 Churn

Churn is a concept that can be borrowed from the marketing literature and characterizes a consumer who frequently changes his provider for another (Glady, Baesens, & Croux, 2009). The concept was first used in the telephone-provider sector, to measure the rate of customers who left a provider over a certain time period. Soon this churn rate became a common measure in subscriber-based service industries, as it provides a good indication of the customer retention rate (Kim, Jung, Suh, & Hwang, 2006). In essence, the concept measures the inflow and outflow of persons or objects from a place or company that they are physically or contractually bound to over a certain time period. If the concept is taken out of its original setting, the churn rate can be applied to the airline sector to explain the ratio between the amount of ceased routes and the total amount of routes in a specific timeframe. If applied this way, the amount of customers that leave the company is replaced by the amount of routes that

(21)

20 are ceased over a specific period. It is also interesting to take into account the net churn, this rate compares the amount of new routes in a certain time period with the amount of ceased routes. A negative net churn would thus mean a decline in the respective route network; vice versa, a positive net churn would indicate a growth of the respective network. As such, a high net churn could be seen as a symptom of an increasingly heavier search for exploitable point-to-point routes in Europe. However, the definition of churn has to be a little more specified to reliably function as a symptom of a heavy search for new routes. An airline can decide to cease a route which has been in operation for many years. Cancelling such a route does not necessarily have to be the result of a route density that was too thin; it can also be the result of for instance changing regulations, the entry of competition, or a fleet redistribution. Therefore, it should be specified in what timeframe a ceased route will be considered as being the result of a too thin route density. In this thesis, routes that are ceased in a time period of 2 year or less after the route has started will be considered as churn. This time period is chosen as it takes some months for an airline before it can correctly assess if a route is profitably enough to stay in operation after the promotional months have passed by. The timeframe of 2 years allows for an operation time of minimal 2 months and a maximum of 36 months. This seems a reasonable time to stay in operation before a final assessment can be made of a route’s profitability.

2.7 Churn measurement

Since churn is a concept not earlier empirically applied to the airline industry, there is neither a pre-defined way to measure churn.The churn rate can be defined in several ways, of which each gives a different perspective on the amount of churn. Three ways are discussed here: The percentual measure, the absolute measure, and the net measure.

First, one can look at the percentual amount of routes that are started and ceased within a period of 2 year. The routes with churn are thereby compared to the total amount of routes in

(22)

21 the network in that particular year. The benefit of this approach is that it provides a clear overview of how many routes there are with churn compared to the total amount of routes that could have had churn. It therefore takes into account the yearly growth or decline of the network size and takes the churn as a percentage of the total routes.

The second approach which can be used to measure the amount of churn is to look at the absolute number of routes that has churn. The benefit of this approach is that it gives a realistic view of the actual increase in routes with churn. However, it does not take into account how much the amount of churn has grown compared to the size of the network.

The third measure is the net approach. In this approach, it is taken into account how many new routes are started in a specific year and how many are ceased again, and the net difference is used to determine the growth or decline of the network size. The disadvantages of this method are that there is less insight in the actual numbers of routes that are started or ceased. Furthermore, it neither says anything about the net growth or decline of the network compared to the actual size of the network. However, it is the most reliable measure when network growth or decline has to be measured.

2.8 Direct and indirect causes of churn

Up to now it was assumed that churn takes place when a route is not profitable enough due to a route density that is too low. For a long time, Ryanair has focused on adding thin routes to expand its network. In the period between 2011 and 2012, Ryanair advertised proudly with 269 route launches (Anna Aero, 2012a). The introduction of new routes from a growing number of operational bases increasingly shows a high trial and error rate. Introduction of a new route leans mostly on free publicity and the offering of very low promotional ticket prices during the first months of operation. When after a few months the ticket prices are raised to a more realistic price level, it will become clear if the minimum occupation rate of

(23)

22 the aircraft can be reached. If this is not the case, the route will be closed again. To illustrate this observation, one can take Ryanair’s operations on Eelde Airport in Groningen, the Netherlands, as an example. In May 2012, the airline decided to start an operating base at this airport, from which three other destinations in Europe would be served. However, in December 2013 Ryanair withdrew its operational base again, arguing this was necessary due to a shortage of airplanes. In the same year Ryanair cut its operations from many more airports – for example from Maastricht airport and Alicante airport. Also, 130 existing routes were ceased (Anna Aero, 2012a). This high amount of cancellations confirms even more the suspicions that Ryanair has problems in finding and maintaining routes that can guarantee an adequate density. It seems that Ryanair is ‘trying out’ potential destinations in the hope to find new city-pairs with a sufficient route density to fill its planes and optimize network profitability. Graham (2013) recognized this trend among LCCs in general, arguing they are footloose in their airport selection and try whichever airport seems to be profitable. When results on a route or operating base are disappointing, the airline seems to simply move its base to another airport where prospects are more hopeful, creating a high churn rate within its network.

However, there could also be more incidental explanations for a route to be ceased. For instance, in 2012 Ryanair was forced to close its operational base at Warshaw Modlin Airport as this airport was closed due to necessary reconstructions of the airstrips. Its operations therefore moved to the nearest airport, Chopin International Airport. When reconstructions were finished the operational base moved back again to Modlin Airport, closing all operations on Chopin International Airport again. Another example of external factors that lead to a shift in operational bases was the close of Baneasa Airport in Bucharest for low-cost traffic, to stimulate growth of the nearby situated Bucharest Otopeni Airport. Ultra low-cost airline

(24)

23 Wizz Air was thereby forced to shift its operational base from this smaller secondary airport to the national airport of Romania.

Such incidental examples only make-up a small part of the churn level, and are largely outside the power and influence of airlines. However, another more indirect reason for route churn can be identified lately. As was mentioned before, Ryanair announced to start focussing on primary airports to escape seasonality problems and to find newly exploitable routes. Recently it started operating bases from primary airports in Brussels, Rome, Milan and Madrid. A shift from secondary to primary airports is expected to be increasingly responsible for the route churn of Ryanair. That is, routes from secondary airports are ceased and moved to new bases on primary airports. This can be illustrated with the following examples of what happened to Bruxelles South Charlerois Airport (BSCA), Girona Airport, Reus Airport and Rome Ciampino. All of them are secondary airports close to the primary hub-airports Bruxelles Zaventem, Barcelona El Prat and Rome Fiumicino and were the home of an operational base of Ryanair. In February 2014 Ryanair started operations from Bruxelles Zaventem, claiming it would not affect its operations from BSCA, and that BSCA would remain its biggest operational base in Northern Europe. However, in the month after the base in Bruxelles Zaventem was opened, BSCA already saw a passenger decrease of 8% (deredactie.be, 2014). If this is a precursor of the rest of the routes at BSCA remains to be seen. Girona Airport and Reus Airport got in a similar situation when an operational base was opened in 2010 from Barcelona El Prat. Both airports saw their passenger figures decrease quickly after, and development of the route network is mainly focused now on Barcelona El Prat (CAPA, 2014d).

2.9 Business model adaptations

The abovementioned shift from secondary to primary airports has to come together with the implementation of a new strategy, as primary airport operations are in contradiction with

(25)

24 lowest-cost principles (Klophaus, Conrady, & Fichert, 2012). Indeed, operations on primary airports come together with higher airport charges and lower turnaround times (Barbot, 2006; de Wit & Zuidberg, 2012). This means that the higher costs of primary airport usage are passed on to customers via higher ticket prices. Higher ticket prices are not in line with Ryanair’s lowest-cost reputation. Furthermore, price increases are also connected with a higher demand for additional service features, something Ryanair did not have its focus on in the past years. Thus, the announcement of primary airport usage seems to demand a shift towards a less ultra low-cost business model.

In addition, the carrier resorted beginning 2014 to a TV marketing campaign in several countries, something it had not done before. Previously the airline only advertised via online banners and sponsored messages in the press. In the TV commercials the airline displays a new passenger-friendly approach with ‘no complications’ as new slogan. This costly marketing campaign is yet another deviation from its lowest-cost approach.

The last significant strategy direction change to mention here is the multi-year deal it signed with a global distribution system (GDS) agency to sell flights through a third party. The objective of this deal is to double the number of business passengers within the next few years, a new passenger group not targeted by the airline before. Ryanair’s chief marketing officer stated that flights which are booked using a third party would not involve additional costs for passengers (Telegraph, 2014), but obviously the deal has to have implications for the cost base of the company.

The above-mentioned developments are interesting examples of Ryanair’s business model adaptations to counter market saturation. However, if these new directions will help them retain network growth remains to be seen. Especially the shift towards primary airports might be a very challenging one, as primary airport operations bring with them a

(26)

25 disadvantaging feature: Direct competition with second-biggest LCC EasyJet, whose business model is configured to operations at primary airports. This means they will create an increasingly overlapping catchment area and will increasingly have to compete for the same target passengers (such as business travellers). Currently, Ryanair and EasyJet have an overlap of 15 primary airports in their network (table 3). However, this amount is expected to grow rapidly the coming years due to Ryanair’s business model changes.

Table 3: Overlapping primary airports between Ryanair and EasyJet Source: OAG database 2014

A development towards overlapping catchment areas will become even more problematic when considering that both airlines have big expansion plans for the future. Graphs 4 and 5 show the orders of both airlines placed at Airbus and Boeing, another factor that is likely to contribute to market saturation.

Graph 4: Aircraft delivery orders EasyJet Graph 5: Aircraft delivery orders Ryanair Source: CAPA (2013c) Source: CAPA (2013d)

Overlapping primary airports

Barcelona International Airport London Gatwick Airport Manchester Airport Brussels Airport London Stansted Airport Nice-Côte d’Azur Airport Eleftherios Venizelos International

Airport (Athens)

Madrid Barajas International Airport Palma de Mallorca Airport

Leonardo da Vinci Fiumicino International Airport (Rome)

Málaga Airport Ruzynĕ International Airport (Prague)

(27)

26 As can be seen, Ryanair has 180 aircraft orders placed at Boeing for the period 2014-2019 (CAPA, 2013d). EasyJet has orders for 147 aircraft at Airbus to be delivered before 2022 (CAPA, 2013c). These orders will partly replace existing aircraft and will partly be used on new destinations. Yet, if Ryanair is struggling already with finding new destinations and maintaining its flight frequency, the orders for new aircraft will impede this search.

All of the developments mentioned above highly suggest a limiting growth for the ULCC business model. The stagnating market in which low density routes provide less and less possibilities for exploitable point-to-point city pairs will lead to a declining growth for these ULCCs. The try-out of these routes will lead to an increasing route churn and instable networks. As a response one sees that business adaptations are taking place such as the shift towards primary airports. However, this brings an increasing competition with established LCC players active on this airport type. This may result in a second-order saturation effect in routes at primary airports as well. Therefore, it remains questionable if market saturation is avoided with this solution. To forecast future network development, it is necessary to know what type of routes have the highest chance of churn. Moreover, only the case of Ryanair has been discussed above, however, other airlines are struggling with route churn as well. To investigate if there are specific differences between carriers with regards to their route churn would tell us something about strategy differences and network focus of the respective airlines.

2.10 Research questions

As is obvious from the previous paragraphs, the focus in this thesis will be on ultra low-cost airline Ryanair. Three other airlines – EasyJet, Norwegian and Wizz Air – will be included in the research as well, to function as benchmarks for Ryanair and to provide additional evidence for the assumption of market saturation. The choice for these airlines is based on the study previously mentioned by Klophaus et al. (2012), who classified low-cost airlines into four

(28)

27 different business model types. The airlines chosen are divided over the first three classifications, which are either archetypical low-cost or hybrid low-cost business models. Of the first type - pure low-cost - 2 airlines are chosen. Ryanair and Wizz Air both fit into this lowest-cost type. Ryanair is chosen as it is the market leader and Wizz Air as it is still a relatively small player, which makes comparison in their behaviours and strategies interesting. EasyJet is chosen as an example of a hybrid low-cost airline with dominant LCC characteristics. EasyJet started in the same period as Ryanair did, but chose a very different business model than the original Southwest strategy, which turned out to make it a highly successful airline. Including this airline thus provides opportunity to see how stable the network of such a business model is and if their churn typology differs significantly from the ULCC model’s churn. The fourth airline included is Norwegian. Norwegian is classified as a hybrid airline with dominant FSC characteristics (Klophaus et al., 2012). This diversifies the sample used and gives additional insight in the characteristics of route churn. Each of these four carriers belong to the first and second tier of European LCCs measured in market size.

In this thesis three research questions will be central. The first question is “does a

saturation of the market lead to an increase in route churn?” This question seeks to find an

answer to the question if the number of routes with churn has increased over the years, and if so, what this means for airlines. The second question refers to the strategy chosen by each airline with regards to their route continuity and discontinuity. It should therefore provide insight in how individual carriers manage their network development. The question is “what

type of routes have the highest chance of being ceased per airline?” The final question

focuses on the development and future perspectives on route churn, and is formulated as follows “will the amount of route churn increase in the next years?”

These three questions will be elaborated in more detail in the research design and hypotheses to be tested, as formulated in section 4.1.

(29)

28

3.0 Business models of the 4 selected LCCs

3.1 Ryanair

Ryanair is the biggest low-cost airline in Europe, with a LCC market share in 2012 of 13% (Financial Times, 2013). It has always been the prime example of a ‘core’ low-cost airline or ultra low-cost carrier, also termed lowest-cost or, in terms of the classification of Klophaus et al. (2012), pure low-cost airline. Ryanair’s strategy has already been largely discussed. Ryanair is known for its use of mainly secondary and tertiary airports, no-frills service, a high seat density and point-to-point operations. Low-cost airlines have to make a trade-off between service elements and costs. With regards to this trade-off, Ryanair performs very low on service elements compared with other LCCs. Nevertheless, passengers see their services as the highest value-for-money due to its considerably lower fares (Lawton, 2002).

In 2014, Ryanair’s fleet consists of 298 Boeing 737-800s, which are configured for 189 economy passengers (Planespotters, 2014). Moreover, its expansion plans are expressed in their orders placed at Boeing in 2013 for an additional 175 new aircraft (Ryanair, 2014d).

Ryanair is often compared with its American counterpart Southwest Airlines. This first successful low-cost airline has always been the prime example for Ryanair and the development of both LCCs have been very alike over the years. Southwest’s development can be split into two phases. In the first phase (from the 1970s to 2006) Southwest operated as a ULCC from and to secondary airports. However, saturation of the US airline market forced the airline to seek for new growth opportunities, thereby starting in 2007 with a shift towards primary airports. Thus, Ryanair was until recently following the first phase of Southwest’s development, but it is adapting its model now to be aligned with the second phase of Southwest. EasyJet’s model, however, had been following this Southwest phase 2 model since the beginning.

(30)

29 3.2 EasyJet

EasyJet is Europe’s second biggest low-cost airline, after Ryanair, with a European market share of 9% in 2012 (Financial Times, 2013). The airline is described by Klophaus et al. (2012) as a hybrid airline with dominating low-cost characteristics. As such, it does not fulfil all cost requirements but incorporates also several full-service elements. In line with low-cost characteristics, EasyJet carries out a philosophy of operating with a mostly homogenous fleet. It used to fly with Boeing 737 aircraft, but since 2002 it switched to Airbus. Currently they fly with over 200 aircraft, of which most of them are Airbus A-319s which have space for 156 passengers per flight. Further, they have orders placed at Airbus for a number of 135 new aircraft, to be fulfilled by 2022 (EasyJet, 2013). Like Ryanair, EasyJet adopted its business model from the American carrier Southwest and adapted it to the European market. However, the strategies of the two biggest European LCCs are not entirely the same. EasyJet’s strategy is to build high frequencies on major city pairs. To achieve this it mainly serves, contrary to Ryanair, primary airports at a higher cost (Lawton, 2002). An incidental consequence of serving primary airports is the presence of a higher yield passenger segment, like business travellers. This customer group is attracted to primary airports by better airport facilities and better infrastructure connection, and is willing to pay a little more for these additional features. By offering additional services like priority boarding, EasyJet is able to attract these price-sensitive higher segment passengers. Moreover, this differentiating strategy helped them until now to avoid direct competition with Ryanair and other lowest-cost airliners.

EasyJet is mostly present in Western Europe, of which the UK, Spain, France, and Italy are its most important countries (Anna Aero, 2012b).

(31)

30 3.3 Norwegian

Norwegian Air Shuttle ASA (from now on: Norwegian) is Norway’s largest low-cost airline and the third biggest ELFAA member in terms of passenger numbers (ELFAA, 2013). Norwegian is described as a hybrid airline with dominating FSC characteristics (Klophaus et al., 2012). Norwegian has a particular strong presence in their home markets in Scandinavia. Their route portfolio is unique for a LCC in that it includes not only Europe, but also several destinations to North-Africa, the Middle East, United States, and Asia. As such, it is the first European low-cost carrier who consistently operates on long-haul destinations outside Europe. Norwegian started with these inter-continental flights as they were one of the first airlines that were able to order the Boeing 787 Dreamliner in a very early stage. These aircraft are relatively light, have lesser fuel costs, and are relatively small. This aircraft type is therefore very suitable for longer-haul low-cost operations which do not require the use of a hub-and-spoke system. A favourable feature of operating outside Europe is the possibility to attract cheaper employees from the operating bases outside Europe. For instance, flights to Asia are mainly performed by Asian employees and flights to the USA by American employees. The attraction of cheap personnel ensures the company’s low-cost operations and a spread of their costs on a global scale. However, these favourable conditions are becoming the subject of an aviation-political debate where the question is raised if these developments start moving towards convenience-flagging such as what happened in the marine transport sector (CAPA, 2014c). These flags of convenience have led to safety problems in the shipping world, but this seems to be unlikely as Norwegian long haul aircraft has been registered in Ireland in conformity with EU safety regulations. Labour unions in the USA as well as European and US network carriers are resisting against the business model of Norwegian as it would give them an unfair advantage and may undermine the wage levels of US airline crews.

(32)

31 Norwegian currently operates with 89 aircraft, of which most are Boeing 737-800s – configured with 186 seats – complemented with some smaller Boeing 737-300s and 3 Boeing 787-8 Dreamliners. In 2012, Norwegian ordered 222 aircraft at Boeing and Airbus, Europe’s largest aircraft order in history (Norwegian, 2014). Of these orders, 14 are for the 787-8 Dreamliner, which indicates plans to further expand their long-haul operations. The orders at Airbus seem to contradict the low-cost philosophy of operating on a homogenous fleet. However, the size of the orders should ensure that scale economies are still possible.

3.4 Wizz Air

Wizz Air is a relatively young Hungarian airline which started operating in 2003. Together with Ryanair and Flybe, Wizz Air is among the only low-cost airlines that stick to the strategy of serving secondary and tertiary airports, therefore it is also classified as a pure low-cost airline (Klophaus et al., 2012). The airline has focused mainly on Central and Eastern Europe where it reached a leading LCC position in 7 countries in 2012 (CAPA,2013e; Anna Aero, 2012c). Its fleet consists of 46 Airbus A320s with a seat capacity of 180 per aircraft. The coming years Wizz Air plans to expand its fleet to 132 aircraft for the year 2017 (Wizz Air, 2014). Wizz Air has the second lowest cost base among European airlines, after Ryanair (graph 6).

With its future expansion plans, Wizz Air will enforce the competition with Ryanair who has a very comparable lowest-cost strategy.

(33)

32

Graph 6: Cost-base European LCCs in 2012 Source: CAPA, 2013

4.0 Model

The literature in the previous section gives an overview of the route density problems in the airline industry and how route churn is related to these problems. This section will start with a discussion of the hypotheses development and end with the conceptual framework.

4.1. Hypotheses

The churn rate will give us an indication of problems and developments within the low-cost airline network, and as such provides complementing evidence to the research of de Wit and Zuidberg (2012) for a saturating market.

Obviously, the rate of churn will vary across airlines and will develop differently. To answer the research question which type of routes have the highest chance on churn, hypotheses will be developed that can help to characterize these type of routes. This characterization will be based on several factors that could influence the churn rate. In this thesis will be argued that churn is influenced by airline strategy, geographical area, airport type, and route lifetime. This section will further elaborate on these four factors.

(34)

33

4.1.1 Churn and network size

In the past years the intensity of competition between carriers has increased. Many routes between new city pairs are launched over the years to broaden the networks and to attract new customers. Over time, it becomes harder to find new routes which can still be exploited profitably. Therefore, even city pairs are tested with a thinner route density and thus less promising expectations. Some routes, however, are not able to attract enough passengers to keep the route profitable. Therewith these low density routes have to be closed again. In this thesis is argued that the level of churn should have grown over the past years due to an increasing network size, as more and more low density routes have been tried. As such, the amount of routes that is ceased within two years after its starting date – the churn rate - increases over the years. The following hypothesis is formulated:

H1: The number of routes with churn has significantly increased as airline network sizes have grown over the years.

4.1.2 Route lifetime

Route churn was defined as the ceasing of a route within two years after the route has opened. However, it also happens that a route is ceased after this time period of two years. In the definition of churn used in this thesis these routes are not accounted for. The reason for a route to be ceased after a time period of 2 years is only indirectly linked to route density problems. A route can be ceased due to increased competition on the airport or on the route itself, a route can be moved towards a primary airport due to strategy adaptations, a route has to be closed due to external factors such as legislations, and many more reasons can be thought of. In general however, the longer a route is in operation, the longer it has proven to be profitable or of strategic importance to an airline. It is thereby expected that a route has lesser chance of being ceased the longer it is in operation. Therefore, the following hypothesis is formulated:

(35)

34

H2: Churn has a specific lifetime, after a certain amount of operation years the chance that a route will be ceased decreases.

4.1.3 Airline strategy

Continuing on the previous hypotheses, it is expected that some airlines have more problems with route churn than others. In specific, Europe’s biggest LCC Ryanair is known for its extensive expansion over the past years. Its strategy has been focused on expanding the amount of routes between city pairs on secondary and tertiary airports to maintain the lowest-cost status. Currently the carrier operates on almost 2500 from 57 bases (Ryanair, 2013b). In the literature review was discussed that Ryanair shows increasingly more signs that finding new exploitable routes becomes harder the more they already have in operation, and more route options fail in the search for exploitable city pairs. Indeed, it is reasonable that Ryanair has approached the maximum amount of exploitable point-to-point destinations within Europe. This is partly explained by the limited possibilities of its current lowest-cost strategy, that predominantly allows operations on secondary or tertiary airports, no connecting hub possibilities, and its low service level. In this regard, expanding the route network might be easier for hybrid airlines like Norwegian or EasyJet. Both airlines provide to a certain degree connection possibilities to other flights at major airports. This provides the opportunity to attract additional customers. They also extend their target group by focusing on the price-sensitive business travellers, thereby generating more demand. It is thus expected that these airlines are not necessarily forced yet to ‘test’ low-density routes to expand their networks, and as such are expected to show less churn in their route network. The following hypothesis is formulated:

H3: Route churn will be higher in the networks of pure low-cost airlines than in the networks of hybrid airlines.

(36)

35

4.1.4 Airport type

The type of airport an airline operates on may also influence the churn rate. In this thesis a distinction is made between primary, secondary and tertiary airports. Many different definitions of these airport types exist. Postorino (2010) argues that an airport can be classified as primary when the number of yearly passengers is bigger than five million per year. However, Murakami (2011) defines a primary airport as the largest airport where FSCs are mainly based within a certain area, regardless of the size of the airport or whether LCCs are present. In general, primary airports are associated with competition from hub airlines, high airport charges, and low turnaround times. Therefore, they are often neglected by lowest-cost airlines to reduce lowest-costs. However, operating on primary airports does have its benefits. Primary airports have lesser route density constraints as their hub functions provide additional opportunities to attract less seasonal and higher passenger volume (de Wit & Zuidberg, 2012). Also, passengers tend to pay more for flights operated from and to primary airports, as these airports have better facilities, are usually closer to dense population areas, and have better and more diverse landside infrastructure connections.

Secondary and tertiary airports are often unsaturated reliever airports (Dobruszkes, 2006; Airline Business Models, 2007), that complement the main primary airport of a city. The locations of these airports are often at a substantial distance from the cities that they claim to serve (Dobruszkes, 2013). In the definition of Klophaus et al. (2012, p. 56), secondary airports are defined as “those not served by the ‘national’ carrier with aircraft having at least 100 seats”. Thus, main carriers only fly on these airport with small aircraft sizes. Secondary and tertiary airports have more problems with route density as their reach is smaller and they tend to miss connection opportunities available at primary hub airports. Due to this problem with route density, secondary and tertiary airports might be more challenging to operate from in terms of attracting enough customers to operate on with a margin. If, after

(37)

36 testing the route from a particular airport, route density tends to be too low, carriers might consider withdrawal from these airports, leading to a higher churn rate from these airport types.

The advantages of primary airport usage are backed up by the shifting strategy seen at Ryanair. The airline tries to include more primary airports within its network, as it believes this provides more opportunities for future growth. This shift to primary airports leads to a new network development for Ryanair. The shift is also paired with ceased routes at secondary and tertiary airports which are moved to the primary airports. Therefore, this strategy adaptation will lead to a higher churn rate at secondary and tertiary airports. Therefore, the following hypotheses are formulated:

H4a: The amount of routes with churn will be higher when a route is operated from a secondary airport type than from a primary airport type.

H4b: The amount of routes with churn will be higher when a route is operated from a tertiary airport type than from a primary airport type.

4.1.5 Geographical area

Besides airline strategy and airport type is there another factor that could influence route churn, namely the geographical location of operations. Northwestern Europe seems to be a logical place for an airline to start operating, considering the fact that it is a high density area with many big cities and high share of important airports (Maertens, 2012). Indeed, the growth of low-cost airlines started mainly in this part of Europe, where Ryanair and EasyJet rapidly expanded their networks from their home bases in the United Kingdom. A route count from the four airlines Ryanair, EasyJet, Norwegian and Wizz Air shows that most of the routes are operated from and/or to Northwestern Europe (5782), followed by routes from

(38)

37 and/or to the Mediterranean (4764) and a much smaller percentage from and/or to Eastern Europe (1883)(OAG database 2014).1

The seasonality issue earlier discussed in chapter 2.4 can also be linked to geographical area. It was already mentioned that the selected airlines are increasingly faced with seasonality issues, being forced to close routes off-season or decrease flight frequencies drastically while no replacement routes are available for them.

Graph 7: Seasonal route churn per year for selected airlines Source: OAG database, 2014

Graph 7 shows the amount of seasonal churn, thereby showing the amount of routes only in operation during the peak-season, and closed during off-season. As can be concluded from this graph, Ryanair is affected disproportionately by this seasonality issue. When considering that 2/3 of their operations is operated from or to the Northwestern part of Europe, one could expect a high level of seasonal churn in this area. In addition, the density problems of Ryanair have been discussed before in which the airline heavily tries to seek low density routes to expand its network with, a search which is increasingly failing. This also adds to the

1_{The counting of this areas is based on how often these regions appear in route combinations. The regions}

counted included routes that were either operated from or had a destination to the specific area. 0 200 400 600 800 1000 N umb er of ro ut es w ith se as on al ity Season

Seasonal route churn per airline

Ryanair EasyJet Norwegian Wizz Air

(39)

38 expectation that routes with churn will be found in the area in which Ryanair is the most active.

With regards to Wizz Air, their business strategy is relatively comparable to that of Ryanair. However, the airline operates on a much smaller scale, and focuses on an area in which the volume of air traffic has just started to grow. Their operating market includes countries with a relatively low GNP per capita. As such, low LCC prices are the only motivation of demand for the airline product, and demand is not generated as a substitution from the FSC to the LCC product like in Northwestern Europe. Since demand in Wizz Air’s operating market is still developing, it is expected that this airline will have less problems with route density in its network than Ryanair and that the level of churn is lower in this airline’s home-market.

In consideration of all the arguments above, the following hypothesis is formulated:

H5: The level of route churn will be higher in the Northwestern part of Europe than in other parts of Europe, i.e. the Mediterranean or Eastern Europe.

4.2 Conceptual framework

The four factors mentioned above all influence the churn rate of a specific airline. Furthermore, these factors are also interdependent. For instance, when an airline is active on an airport in or a route to a less dense geographical area, other players will not easily be attracted as catchment areas will not be sufficient enough to serve more players. The hypotheses stated are depicted in the following conceptual model (figure 1).

(40)

39

Figure 1: Conceptual model

5.0 Methodology

This section describes the methodology used to test the hypotheses developed in the previous section. First, the data collection and sample is discussed, followed by a description of the dependent and independent variables. This section ends with a description of the empirical approach.

5.1 Data collection

The AOG database was used to extract data about airlines for the twelve year period 2002/2003 – 2013/2014. The sample consists of the four airlines Ryanair, EasyJet, Norwegian and Wizz Air, and contains information on all operated routes of these airlines in the given timeframe. Within the database, 8052 routes are included. A list of airlines and the number of routes they operated over the twelve year timeframe can be found below.

In order to apply valid tests with the database, all cases were excluded that did not have a starting year in the period between 2002 and 2014 (n=236), as no conclusions can be drawn about the operation time of the cases starting before 2002 or still being served after 2014. Also, all cases were excluded that were ceased during the timeframe but restarted again (n=330). This was decided after it was too disputable if these routes could be characterized as churn routes, and as it was very hard to include these routes in measurements where operation

Churn Network size Airline strategy Airport type Geographical region Route lifetime

The end of growth for low-cost airline networks? : route churn as measure for route density problems in airline networks