Evolution of Service Products on Travel Sites and the Effect of that on Traffic: A Longitudinal Study

The Internet has started a process of change in the travel product buying habits of both leisure and business customers. The Internet has allowed people to access travel information with minimum inconvenience. In addition it gave the possibility to purchase travel products online (Buhalis & Licata, 2002). Tourism related services are among the leading services to be promoted and distributed via the Internet (Sussman & Baker, 1996, Millman, 1998). More and more travel companies are starting to develop web presence and offer their services online. As firms gain experience with internet technologies, their web sites move from a static presence through increasing levels of interactivity to dynamic sites (Poon & Swatson, 1999). Web presence evolves over time. Services offered over the web have become sophisticated in many countries. This study aims to determine the evolution of service provision by Egyptian travel sites and how these services change in terms of their level of sophistication over the period of the study which is ten years. This content analysis consists of a small scale survey done on 30 Egyptian travel agents’ web sites that attempt to examine whether Egyptian travel websites are static or dynamic in terms of the services that they provide and whether they provide simple or sophisticated travel services. Additionally, the study attempts to determine the popularity of these websites through the number of links to these sites. Links maybe viewed as the equivalent of a referral or word of mouth (Ennew et al., 2005). Both popularity and the nature of the services provided by these websites will be used to determine the traffic on these sites. In examining the nature of services provided, the website itself is viewed as an overall service offering that is composed of different travel products and services. In order to determine the level of sophistication of these travel sites, the nature and composition of products and services offered by these sites were first examined. In classifying the different types of services available on the Egyptian travel agents’ web sites, a framework adapted from Kotler (1997) “Five levels of a product” was developed. The main contribution of this study is a theoretical one whereby a deep analysis of the various models explaining the levels of a product was conducted, comparisons were made and a mapping of the four models outlined was done. Additionally, adapting Kotler’s model of levels of a product and applying it to the travel service are main contributions here. Additionally, measuring evolution over a period of ten years is another main contribution in this study that does not take a cross section as most studies do. The study began by identifying the Egyptian travel agents that have online presence and that provide online travel services. The target group for this study consists of companies that do inbound tourism and their focus is to attract tourists to the country. Asearch for Egyptian travel agents who own a website was done via metacrawler search engine which identified 100 web sites. When these web sites were first examined in June 2004 it was found that some of these web sites were passive pages that only included contact information of the travel agent with no services provided at all. Other web sites included small pages that were not actually used since they only included general information on Egypt with again no specific services promoted or offered. Out of these 100 web sites, only thirty travel agents were identified, based on the researcher’s judgment, as usable ones. These were identified as usable if they actually provided or promoted online travel services such as tour packages, travel information, hotel information, etc. and not just general information about the company. These thirty travel agents’ sites identified from sites of tourist organizations including the Association of Egyptian Travel Business on the Internet (AETBI) represent the sample of this study. Four rounds of data collection were conducted over a period of 10 years. Two rounds of data collection were made in 2004 and two rounds were made in 2014. Data from the travel agents’ sites were collected over a two www.ccsenet.org/ijms International Journal of Marketing Studies Vol. 7, No. 3; 2015 28 weeks period in each of the four rounds. These four rounds provided sufficient data on where do these web sites stand in terms of the features they provide and the evolution of these features over the period of the study. In conjunction to collecting data on features of web sites, data was also collected on the popularity of these web sites through a software program called Alexa that showed the traffic rank and number of links of each site. Alexa is a page counter site with one of the widest reaches in terms of sites covered and number of users. Data on traffic rank and links was collected over one day in each round. Results provide clear concrete evidence that websites’ features and links are significant determinants of traffic generated on these websites. Descriptive analysis was conducted to determine the ratio of the different categories of services provided as well as the change in traffic and the number of links over the period of the study which is 10 years. Results indicate that as companies moved from having simple websites with basic travel information to being more interactive, the number of visitors illustrated by traffic and the popularity of those sites increase as shown by the number of links. The results of this content analysis study taps on an unexplored area and provide useful insights for marketers on how they can generate more traffic to their websites by focusing on developing a distinctive content on these sites and also by focusing on the visibility of their sites thus enhancing the popularity or links to their sites (Dreze & Zufryden, 2004). Results also show that travel companies use the web much more for promotion rather than for distribution since most travel agents are using it basically for information provision. This is in line with literature whereby perceived risk was consistently found to have a negative relationship with Web adoption and use (Lockett & Littler, 1997; Dohetry et al., 2003). It might be that the majority of travel sites are used for information provision and communication only rather than distribution out of a fear of the risks associated with conducting a full transaction and payment online.

weeks period in each of the four rounds.These four rounds provided sufficient data on where do these web sites stand in terms of the features they provide and the evolution of these features over the period of the study.In conjunction to collecting data on features of web sites, data was also collected on the popularity of these web sites through a software program called Alexa that showed the traffic rank and number of links of each site.Alexa is a page counter site with one of the widest reaches in terms of sites covered and number of users.Data on traffic rank and links was collected over one day in each round.Results provide clear concrete evidence that websites' features and links are significant determinants of traffic generated on these websites.Descriptive analysis was conducted to determine the ratio of the different categories of services provided as well as the change in traffic and the number of links over the period of the study which is 10 years.Results indicate that as companies moved from having simple websites with basic travel information to being more interactive, the number of visitors illustrated by traffic and the popularity of those sites increase as shown by the number of links.The results of this content analysis study taps on an unexplored area and provide useful insights for marketers on how they can generate more traffic to their websites by focusing on developing a distinctive content on these sites and also by focusing on the visibility of their sites thus enhancing the popularity or links to their sites (Dreze & Zufryden, 2004).
Results also show that travel companies use the web much more for promotion rather than for distribution since most travel agents are using it basically for information provision.This is in line with literature whereby perceived risk was consistently found to have a negative relationship with Web adoption and use (Lockett & Littler, 1997;Dohetry et al., 2003).It might be that the majority of travel sites are used for information provision and communication only rather than distribution out of a fear of the risks associated with conducting a full transaction and payment online.
Keywords: travel websites, traffic, evolution of services, content analysis 1. Background

Alternative Models Explaining the Composition of a Service
In order to categorize the different types of services available on Egyptian travel agents' sites, it is important to examine the different existing models describing the nature and composition of services.A review of existing models shows that there have been some attempts made to explain the composition of services by examining the services offered and putting them under different levels/categories according to the purpose they satisfy (Kotler, 1997;Gronroos, 1990Gronroos, :2000;;Storey & Easingwood, 1998;Fitzsimmons & Fitzsimmons, 1994).An additional model (Davidoff et al., 2000) that has focused especially on web sites is also being considered.It is important to mention that among the four models, Kotler's (1997) classification offers the most comprehensive framework since it includes precisely defined levels with clear distinctions between levels in terms of the purpose served and the value added.Additionally, it includes more levels than the other three models and thus gives more latitude for classifying services.Moreover, the clear distinction among the different levels under Kotler's framework does not lead to confusion in classifying different services as might be the case with some of the other models as will be explained below.
It is worth noting that none of these four models discuss how services evolve.They are descriptive static models that explain the different layers of a service but not the evolution of the service over time.They describe service elements that add to customer value.Nevertheless, they provide us with an adequate base to describe services now and in the future by the different levels of a service that might be offered.We can thus use these models to compare between services over time which serves the purpose of this study.The following paragraphs will provide an overview of each of these models and explain the similarities and differences among them in describing the different levels of a service.
1.1.1Kotler's Five Levels of a Product Kotler (1997) developed a model of levels of a product based on the idea that a product or service can classified under five different levels from core to potential according to the benefit it provides.Kotler (1997, p.430) defined a product as "anything that can be offered to a market to satisfy a want or need."Thus the term product is used to refer to both physical goods as well as intangible services.There are five levels of a product that the marketer should think about and that provide more value at each level (Kotler, 1997).In the following paragraph the researcher will provide a brief explanation of how Kotler defined the five levels of a product. Core benefit; is the fundamental benefit or service that the customer is actually buying.
 Basic product; is the tangible product.This includes all features that describe a product including colour, brand, package, features and so on. Expected product; includes a set of attributes that the buyer normally expects and that constitute the buyer's minimum expectation in which not much preference is shown. Augmented product; includes additional benefits and services that distinguish a company's product from its competitors. Potential product; is the value added in terms of new ways to satisfy a customer and further distinguish the company's product/offer.The aim behind potential products is to delight customers by exceeding their expectations.(Kotler, 1997, p. 431/2).
These different levels of a product have been described in a generic sense by Kotler and thus they can be applied to services.It is important to recognize that this model focuses on providing more customer value at each level of product/service by adding more features that evolve from satisfying basic requirements to meeting normal expectations to exceeding expectations and even surprising customers.Thus, if used as the framework for classifying services provided by Egyptian travel agents' web sites, the classification should be based on what benefits each category of services provides to tourists and whether these satisfy basic, expected or above expected needs.It is important to mention that Kolter's framework is a well-established one that provides a comprehensive leveling of the different products serving different customers'needs.
1.1.2Gronroos's Basic Service Package Gronroos (1990Gronroos ( :2000) ) developed a model of service features based on the idea that any service is composed of a bundle of different services, tangibles and intangibles, which together form the service.This bundle of services, according to him, is the "basic service package" that consists of three levels of services: core, facilitating and supporting services.Core services constitute the main reason for a firm's existence.A hotel, for example, provides a lodging service and without this service the hotel cannot exist.The lodging service is regarded as a core service in this regard.In order to support this existence another group of services, the facilitating services, is required.Such services facilitate the consumption of the core services that cannot be consumed otherwise.A hotel cannot function without a reception service.The reception service is needed to facilitate the provision of the core service which is the lodging service.A third group of services, the supporting services, is used to differentiate the service from its competitors by adding value to it.Hotel restaurants and entertainment facilities are examples of supporting services.A hotel can differentiate itself by providing distinguished restaurant atmospheres, good quality food, a large variety of entertainment facilities etc. Supporting services are not necessary for the consumption of the core services, they are a way to distinguish a service provider's services from the competition.Gronroos (1990Gronroos ( :2000) ) explains that it is not always easy to distinguish between a facilitating service and a supporting service in all situations.A service that may be considered facilitating in one situation, for example an in-flight meal on a long destination route maybe considered a supporting service on a short flight.
1.1.3Davidoff et al. Model of Cybermediary Functions Davidoff et al. (2000) developed a model that utilizes the same idea of different layers of a service but with a special emphasis on the internet.It seeks to classify web sites based on the types of services they provide from basic to advanced functions.It explains the basic and advanced functions that cybermediaries are expected to perform in order to create value and be successful on the web.It discusses four levels of cybermediary functions being; basic, basic commercial, advanced and advanced commercial.Basic functions include facilitating information exchange, aggregating buyers and suppliers demands and matching their needs.By carrying out these functions, cybermediaries reduce customers' search costs, provide information and provide a mechanism for exchange.Moreover, they combine buyer and supplier demands to achieve economies of scale and scope and introduce potential buyers to suppliers to match needs.Basic commercial functions include conducting e-commerce by enabling online commercial transactions and thus generating revenue.The last two levels deal with functions such as building trust, providing a sense of community, authenticating and hence leading to the most advanced level of value capture through personalizing the virtual experience.This model is important in that it examines functions performed by web sites, which is unlike most studies that focused only on classifying websites based on factors such as navigation, speed, freshness and graphics in determining web sites' success.However, this model can be addressed more toward web sites that perform a hierarchy of functions and thus actually provide different levels of services that can be classified according to the levels of cybermediary functions proposed..This is quite unachievable by Egyptian travel agents web sites since most of these sites perform basic functions that fall under information provision with few ones focusing on more advanced functions.Although a description of the different levels may be context specific such that what could be defined as a basic function in one context could be considered as advanced in another context, it will still be difficult to classify the services provided by Egyptian travel agents'sites based on this framework as the majority of services would fall under basic cybermediary functions.
Results of surveying Egyptian travel agents sites show that the majority of the sites provide basic cybermediary functions with only a few that attempt to provide value through personalizing the virtual experience as proposed by Davidoff et al. (2000).According to Breitenbach and Van Doren (1998) the challenge for e-marketers is to provide users with an experience that is not easily replicated by conventional media.Web sites should provide added value to visitors to encourage them to visit and to retain them.
1.1.4Fitzsimmons and Fitzsimmons Model of Service Package Fitzsimmons and Fitzsimmons (1994) developed a framework that describes the composition of a service in terms of four features that, according to them, are experienced by the customer and form the basis of his or her perception of the service.These features form "the service package" that consists of a bundle of goods and services that include the following:  Supporting facility; physical resources necessary to provide a service.A hotel is an example of a supporting facility. Facilitating goods; the materials purchased or consumed by buyers.Soap and paper in a hotel are examples of facilitating goods. Explicit services; benefits that are observable by the senses and form the essential features of the service.A clean room in a hotel is an example of an explicit service. Implicit services; psychological benefits that consumers may sense indirectly.A friendly hotel receptionist is an example of an implicit service.Fitzsimmons and Fitzsimmons (1994) explain that all these four features are complementary and form the basis of customers'perceptions .This model cannot be easily applied to travel agents'web sites as it relies heavily in its four features on the physical evidence in the customer's experience.Thus, although this model attempts to discuss the attributes of a service package, it cannot be applied to web sites since it concentrates on physical rather than virtual attributes.

Comparing the Four Models
In deciding on which model to use as a basis for this study, comparisons were made between the four models in terms of purpose and the levels defined by each.Table (1) below and the following discussion provide an explanation of the similarities and differences between these models.In attempting to compare the four models, the model developed by Davidoff et al. (2000) was found to be not easily comparable to the other three models due to fact that it is very specific to web sites whereas the others are more generic.In addition, the distinction between its different levels, especially with regard to comparing basic cybermediary with basic commercial functions as well as advanced cybermediary with advanced commercial functions is not very clear.
Kotler's core benefit corresponds directly to Gronroos core service and matches with Fitzsimmons and Fitzsimmons supporting facility.All three discuss the fundamental benefit or service that is being delivered and that is the reason behind the firm's existence.A travel agent exists to provide a travel service which constitutes the core service of a travel agent.Moving on to the next level, Kotler, Gronroos and Fitzsimmons and Fitzsimmons similarly discuss the tangible/physical elements of a service that are necessary to consume a service.Beds, reception services, soap and paper are all examples of basic or facilitating products to a hotel service.Expected services and explicit services similarly constitute the normal expectations of the buyer with regards to the service.
A clean room in a hotel is an example of such a service.What Gronroos describes as a facilitating service corresponds to both basic and expected services in Kotler's model.Similarly, Gronroos's facilitating services correspond to both facilitating and explicit services in Fitzsimmons and Fitzsimmons model.This is because Gronroos discusses that these are the services needed to facilitate the consumption of the core service and without which the core service will collapse.Thus, implicit within this level are the basic services needed to consume the core service as well as satisfy the customer's normal expectations.However, Gronroos does not provide a clear distinction between basic and expected services.
The levels in Davidoff et al. model that could be seen as corresponding to basic cybermediary functions and basic commercial functions as shown in table(1).This is because they define these two levels to include the basic functions that a cybermediary must perform to acheive success on the web.These include facilitating information exchange, aggregating demands of buyers and suppliers and providing a method of security and authentication of their offering.Thus, these functions could be viewed to constitute the core of existence of the web site and at the same time they form customers'normal expectations.
Moving on to the last two levels in Kotler's model which are augmented and potential services, these correspond to supporting services by Gronroos and implicit services by Fitzsimmons and Fitzsimmons since all deal with layers of a service that add more customer value and exceed customers'expectations.It is clear however, that only Kotler's model provides a clear distinction between the different layers of a service that deal with augmenting the service and exceeding customers'expectations by distinguishing between augmented and potential services.Davidoff's advanced cybermediary and advanced commercial functions match the augmented and potential services by Kotler, supporting services by Gronroos and implicit services by Fitzsimmons and Fitzsimmons because they all deal with the same idea of adding more value to the web site by creating virtual experience and personalizing the interaction.However, there is no clear cut distinction between the purpose of each of these two levels under Davidoff et al. model and thus it is thought that both levels together correspond to Kotler's augmented and potential services.
Travel agents'web site services that could be categorized as augmented and potential services under Kotler's framework can be viewed as supporting services under Gronroos's model, implicit services under Fitzsimmons and Fitzsimmons and advanced cyberrmediary&advanced commercial functions under Davidoff et al.'s model.Virtual tours, for example can be classified as potential services because they go beyond customers' expectations by allowing them to see and judge the place without actually traveling.Therefore, they help decrease perceived risk and make customers more comfortable about their travel choices.Virtual tours, on the other hand, according to Gronroos (1990Gronroos ( :2000) ) can function as a supporting service to hotels, cruises and safaris.Travel agents providing virtual tours can differentiate their offerings from competitors as well as add value to their core services by virtually illustrating the distinctive features of their services to customers and allowing them to make more informed decisions concerning their travel.Based on the same idea of differentiating the company's offering and adding more value to the customer, virtual tours could be defined as advanced cybermediary and advanced commercial functions under Davidoff et al. model.It is difficult, however, to classify virtual tours as implicit services as proposed by Fitzsimmons and Fitzsimmons as their model is based on the physical existence of the customer and implicit services are value added services that are perceived by the customer in his/her interaction with the service provider.
A few travel sites attempt to provide the advanced cybermediary functions proposed by Davidoff et al. (2000) by offering services that would attract customers and lead to more interactivity.One web site, for example, provides a chat room where web site visitors from different parts of the world can chat together and discuss their travel experiences.This service helps build a sense of community whereby people with common interests can interact and exchange travel information.This service also provides incremental value to the web site and differentiates it from its competitors.
Building on the same idea of interactivity, other sites provide an online guest book where travellers can sign in their comments about their travel experiences.Also web sites visitors can provide comments on any aspect of the web site in the guest book.The idea of a guest book is an interesting idea in itself since visitors to web sites can learn more about what other people, who have already travelled with the travel agent, think of services provided by the travel agent and about the places they have visited in Egypt.Sharing other travellers' experiences can help reduce uncertainty about the place and make new visitors more comfortable about their travel choices.Virtual tours can also be regarded as a form of interactivity as they provide the site visitor with virtual experience of the destination.
Other sites attempt to conduct e-commerce transactions by providing the customers with an online marketplace where tourism products can be bought electronically.These sites, however, do not provide the necessary services to secure these online transactions such as money back guarantee or alliances with security organizations.Thus, these sites do not provide the necessary infrastructure for customers to complete their online transactions.Other sites attempt to provide visitors with interesting services that would help capture their attention such as newsletters, e-cards and new coming events.The aim of these services is to create more value by providing attractive services to visitors to capture their attention and retain them in the long term.
Having reviewed the alternative frameworks of service products, it is obvious that there are no major differences between the models.All of them serve the purpose of outlining the different levels of services that could be provided by the supplier or intermediary.However, Kotler (1997) model of "five levels of a product" is the most established and most comprehensive one to be used as a base for this research.This is because this model provides a clear distinction between different levels of product and at the same time provides a variety of levels that provide latitude for choosing which services fall under which category.

The Survey Framework
This study began by identifying Egyptian travel agents that have online presence and that actually provide online travel services.The target group for this study consists of companies that do inbound tourism and bring tourists to the Egypt.A search for Egyptian travel agents who own a website was done via metacrawler search engine that identified 100 web sites.When these sites were first examined in June 2004, it was found that some of them were passive pages that only included contact information of the travel agent and nothing more.Other web sites included few pages that were not actually used since they only included general information on Egypt with no specific services promoted or offered.Out of these 100 web sites, only thirty travel agents were identified, based on the researcher's judgment, as usable ones.These were identified as usable if they actually provided or marketed for online travel services such as tour packages, travel information, hotel information, etc. and not just provided general information about the company.These thirty travel agents' sites identified from sites of tourist organizations including the Association of Egyptian Travel Business on the Internet (AETBI) represent the sample of this study.
In surveying these web sites, the first step was to count the different tourism services provided online by the thirty travel agents in order to prepare a complete list of all services provided by all thirty sites.A list of thirty two services resulted (see Appendix 1).In order to classify these services according to Kotler's framework that is described above, one adaptation was necessary which was to use four rather than five levels of a product.The reason for this change is that the services provided by the travel agents provide very little, almost negligible differentiation between the first two levels of a product in Kotler's framework, being core and basic products.It was found reasonable to merge both the core and basic products into one level that includes both, since most of the services provided by travel agents in this category satisfy both purposes.A hotel as a tourism product for instance, satisfies the core benefit of providing rest and sleep to customers as well as the basic product benefit which are the features that characterize the physical product such as the hotel name, rooms, reception, information desk and so on.
While surveying the websites, distinctions were made among the various services offered online by the travel agents.It was found that the majority of web sites provide tour packages, Nile cruise packages, information about hotels available and so on.These offerings are basic for the travel agents existence because a travel agent cannot exist without offering tour packages and information about the accommodation available at a destination.Thus it was considered that the first level of the model would include these basic offerings that most travel agents provide.Moving a step further was another set of services provided by a large number of web sites but not the majority.These services aim to meet customers' travel needs and thus include elements like travel tips, maps, information on prices and so on.Such services are not vital for the main function of a travel agent but they still meet customers' normal expectations.These offerings could form the level of expected services.
Fewer sites aimed to provide more differentiated offerings than competitors such as tailor-made packages to meet customers' specific travel preferences, providing them with currency converter, weather forecast, extra tour options and so on.These services form a third layer of augmented services.The final set of services included offerings that exceed customers'expectations and aim to delight them.These are provided by very few travel agents and they include services such as online marketplace where they can buy "Ancient Egypt" products, chat rooms where travellers can share their travel experiences, virtual tours demonstrating the actual destinations online and so on.These services form a layer of potential services.
Therefore, a four level model was used to capture the different layers of services provided by travel agents.The following is an explanation of the four levels model adapted by the researcher from Kotler's framework as well as the attributes depicted under each .In order to categorize the services offered by the travel agents' web sites under the four levels model described above, the services provided by each site were thoroughly investigated to determine the frequency of the provision of each service among all the thirty sites studied.Categorizing the attributes was thus based on both the frequency of the provision of each service as well as the author's personal judgment.The following is an explanation of how these attributes were categorized under each level.

Core/Basic Services
These are tangible tourism products that satisfy the main reason of existence of a travel agent that is to provide a travel service.These products/services are regarded as the backbone of a travel agent's existence .These include programs covering the main tourism attractions available in Egypt such as the red sea, the cultural heritage and the desert.It also includes information on accommodation available at destinations since such information is basic for travelling purposes.Included under this category are the following attributes: tour packages, Nile cruise programs, diving programs, excursions, safari and hotels' information.A review of the web sites under study indicated that these services were provided by most of them.

Expected Services
These are expected from a travel agent but are not necessary for its existence, i.e. it can still function without providing them.These meet the customers' normal expectations as to what a travel agent should offer.Services such as providing the traveller with travel tips, maps, prices of destinations, hints on Egypt all satisfy customers' normal expectations and are all provided by a large number of travel agents.Hotel booking and online reservations also form a normal expectation by travellers who need to be able to book online.It is worth noting however, that being part of an emerging economy where IT is not yet advanced and fully secure measures are not yet in place, not all of the sites that offer online hotel booking and reservations automatically provide the traveller with immediate online reservation.Most of these sites ask customers to enter their preferences regarding the accommodation and the duration of travel in a preset form or via email and then provide them with a reply at a later stage.This is basically because there is no adequate level of integration on the B2B level between the different travel parties involved, i.e. travel agents and hotels in this case.One final attribute that is surprisingly provided by a few travel agents is information on restaurants available at destinations.This constitutes a normal expectation by travellers since it is one of the things that they would be interested to know and may consider in making his/her travel choices.Thus information on restaurants was also included among the expected services.

Augmented Services
These are services provided by travel agents that go beyond customers' normal expectations.These are used to distinguish a travel agent's offerings from competitors.Included under this level are all packages that serve customers' special needs such as tailor made packages, hot deals/special offers, extra tour options and honeymoon packages.Also included are attributes that provide travellers with less common features such currency converter, weather forecast and an online guest book where they can record their comments and share their experiences with people from all over the world.Finally, three more attributes were included under this level that are transfers and car rentals, train reservation and flight booking.Web sites that are at the maturity level in terms of web use provide these services as expected services since securing the transportation for the traveller is an important part of any travel and thus should be a normal expectation from a travel agent.In Egypt, however, travel agents' web sites can be considered at adevelopment stage and at the same time there are few strategic alliances formed between the different tourism organizations such as travel agents and transportation companies, thus such services are not normally provided.Travel agents that provide online transfers and car rentals, train reservation or flight booking have gone a step ahead of the normally expected services provided by Egyptian travel agents' web sites and that is the main reason why these services were categorized as augmented.

Potential Services
These are services that are rarely provided by travel agents.They are unusual services that aim to exceed customers' expectations by delighting them.Attributes included here in particular allowed more interaction with travellers.For example, an online marketplace where customers can buy products is provided by few sites, a chat room where travellers can share their travel experiences with people from all over the world, new events are also posted by few sites, Egypt's directory and a newsletter are also provided.Virtual tours, inquiry forms and e-cards were also seen as relevant to this level as they are regarded as extra services.The following section explains in details the approach taken to analyze the websites.

Website Analysis
The purpose of this content analysis is to describe the way the web is being used by travel agents through the service attributes provided on their web sites and explain how these attributes evolved over the period of the study in order to understand the prevailing pattern of activity.Additionally, this evolution is linked to traffic and popularity of these sites demonstrated by the number of links.In order to achieve this objective a systematic analysis of websiteswas needed.A simple descriptive analysis follows to draw conclusions.This section explains the approach taken to analyze the features present on travel agents websites and their evolution over the period of the study which is ten years.
Website analysis started by a count of all the different features present on the thirty chosen travel agents' websites in June 2004.All thirty sites chosen had at least one or more of the facilities needed for travel inquiry, communication, reservation or booking.The counting process of features on all 30 websites resulted in 32 features in total being identified.This was done by visiting each site and counting the number of features present on it.An aggregation of the different features provided by all 30 sites resulted in a total of 32 features.
In order to classify these features under Kotler's framework, two approaches were used.First, the frequency of provision of each service in all 30 sites was calculated.A count of how many sites provided this features was done.Tour packages were provided by all 30 sites in all rounds of data collection.Nile cruise programs were provided by 26 out of the 30 travel sites.The frequency of provision of the feature gave an indication of the importance of the service to the travel agent.Second, the researcher relied on personal judgement to determine which feature should fall under which category.Tour packages for example were seen as essential for the travel agent's existence as they represent the main service the travel agent exists to provide and thus were classified under core/basic services.Virtual tours were seen as potential services as they exceed customers' expectations by providing them with a view of the place they are going to visit, and are provided by very few travel agents (only 2 sites in the second round provided virtual tours).Thus, together, the frequency analysis and the researcher's personal judgement were used to classify the features under kotler's framework.
Four rounds of data collection were conducted over a period of 10 years in order to determine the attributes that each web site provides and the change in these services across the four rounds.Two rounds were conducted in 2004 and another two rounds were conducted in 2014.The time difference between the two rounds in 2004 was six months and the same for the 2014 rounds.Data from the travel agents' sites were collected over a two weeks period in each of the four rounds.These four rounds provided sufficient information on the status of these websites in terms of the features they provide and also on the evolution of these features over the period of the study.Table (2) below shows services provided on travel agents sites in one of the four rounds demonstrated in ratios and percentages.The figures were calculated based on the number of attributes in each level to the total number of attributes of that level for each web site.In conjunction to collecting data on features of web sites, data was also collected on the popularity of these web sites through a software program called Alexa that showed the traffic rank and number of links of each site.Data on traffic rank and links was collected over one day in each round.Data on traffic gives an indication as to whether more customers are willing to visit the websites because of the more sophisticated features provided on it.Links give an indication of the popularity of the website in terms of how visible this website is and this also is related to the features and servicesprovided on the website.This study attempted to make a relationship between the features provided on websites and the evolution of those features and traffic as well as popularity of those sites.The following presents the descriptive analysis conducted on the data and the results of the survey.

Results of the Survey on Travel Agents' Websites
The following table provides a descriptive analysis of the data gathered on services features provided by the 30 Egyptian travel agents' websites over a period of 10 years in order to see the evolution of services provided by these sites.As can be seen in table (3), core services exhibit the highest mean in the four groups of services indicating that these websites focus mainly on providing the services that are basic for a travel agent's existence.Following that, expected services also show a relatively high mean indicating that these services are also adequately provided by travel agents in comparison to the other services.The mean of augmented services is low indicating that these services are not provided by a large number of travel agents.Finally, potential services have the lowest mean indicating that only a few web sites provide these services.This suggests that most travel agents are simple users of the internet whereby their main focus is to provide the basic services necessary for the existence of the travel agent and that target customers' normal expectations.Additionally, fewer companies can be categorized as sophisticated users who aim to provide augmented or potential services that go beyond customers' expectations.
Comparing the four rounds together over the period of the study shows that there has been an increase in the range of services provided over the period of the study although this change is small.The means of the four groups of services across all rounds indicate that few changes have occurred in terms of the services provided on these websites.The core, expected and augmented services show few changes in means between first and fourth rounds while potential services means were the same for all four rounds.Thus in terms of evolution of services over the period of the study, some change has occurred but this change was small indicating that besides providing basic services, these travel sites are quite static in terms of the services they provide on their sites.
If we look at the evolution of all four groups of services taken together over the ten years, indicated by the total services mean, we would notice the same results suggesting that there has been a small increase in the total services provided indicating that these websites have made few changes in terms of features provided on their websites over the period of the study.The total services means, which represent the mix of different service types these firms provide, indicate that only few changes have occurred in terms of services provided over these sites over the ten years.These are in core, expected and augmented services as explained above.This further confirms that in terms of features provided on their websites, these companies are simple users of the web As mentioned earlier, this study attempts to see whether links of these sites changed over the period of the study in order to determine the change in the popularity of these web sites.As can be seen in table (4), the number of links on Travel agents sites across all rounds is small with a mean of around thirty five links for the majority of the sites.One or two sites might have no links in each round and one or two may have very high number shown by the maximum number of links.This wide variation in the number of links among the sites is manifested by the large standard deviation values in all rounds.However, the vast majority of the sites have mean number of links of around thirties.These figures show that Egyptian travel websites have low popularity because of the small number of links that they have.There are some sites that are not even linked to other sites and thus have zero links.This might indicate that Egyptian travel agents need to place more emphasis on marketing their travel sites by linking themselves to popular travel sites in order to achieve global access and exposure.This would help them be visible by a large number of customers who search these popular sites for different purposes.This might also indicate that Egyptian travel agents need to develop their marketing capabilities in order to be able to properly market their travel services on these popular websites that they link themselves too and thus can attract more customers.
As for traffic which is an indicator of the number of visitors on these websites, table (5) shows that there has been a good increase in the number of traffic from round 1 to round 4. Additionally, the mean number of traffic in each round indicates a reasonable number of visitors on each of these sites.It is expected that if these sites work more on developing the services they offer over the internet and move from providing only basic service to providing more augmented and potential services, that this will have a positive impact on traffic on these sites.

Conclusion
The purpose of this research is to study the evolution of online travel services through a survey conducted on a sample of thirty Egyptian travel agents' websites over a period of ten years.The main contribution of this study is a theoretical one.An in depth analysis of the various models explaining the levels of a product was conducted, comparisons were made and a mapping of the four models was outlined .Additionally, adapting Kotler's model of levels of a product and applying it to the travel service is another main contribution in this study.Moreover, measuring evolution over a period of ten years is another main contribution in this study that does not take a cross section but studies change over time.Results of this study show that Egyptian travel agents' websites are quite static in terms of the content provided and that changes do not take place at a fast pace.Additionally, the majority of these sites can be categorized as simple users of the internet who focus to a great extent on the provision of core and expected services rather than augmented and potential services.The sites that provide augmented and potential services are very few indicating that the level of sophistication of these sites is low.This matter might indicate that travel agents need to develop their marketing capabilities in order to be able to provide travellers with good services.Links and traffic also show that athough there has been an increase in the number of links and traffic on these sites from round 1 to round 14, this increase is a modest one which further indicates that more emphasis is needed on developing the content and level of sophistication of the services provided on those sites.

Limitations and Future Research
This study focused on only one sector which is the travel sector.Although this is one of the sectors that has been growing very fast on the internet, it would be interesting to consider other sectors and see how they would compare in terms of the services offered online as well as the evolution of those services and how this affects traffic and links.Another limitation here is the fact that this research work focused on one context which is the emerging country context.More insights could be gained if the same model is applied across countries.This will also provide useful information and comparison across countries.The main contribution of this research is a theoretical one that is based on compiling the existing model of levels of a product and adapting Kotler's model and applying it to a specific industry and context.The research conducted descriptive analysis of the results.Future research can focus on delving more deep into measuring the effect of features and links on traffic.

Table 1 .
Mapping the four models

Table 2 .
Ratios and percentages of service provision by travel agents in round 1

Table 3 .
Mean number of features in each category across all rounds

Table 4 .
Table (4)shows a descriptive analysis of the number of links for the 30 sites across all rounds.Table (5)shows the descriptive analysis of the traffic on the 30 sites across all rounds.Descriptive analysis of the number of links to the 30 websites across all rounds

Table 5 .
Descriptive analysis of traffic on the 30 websites across all rounds