The Internet Travel Industry:

What Consumers Should Expect and Need to Know, and Options for a Better Marketplace {Full Report)

TABLE OF CONTENTS

Executive Summary

The online travel market continues to evolve, creating both risks and opportunities for shoppers. Since Consumer Reports Travel Letter (CRTL) conducted its assessment of travel Web sites in October 2000, many changes, both good and bad, have occurred. While a wide variety of airline ticket-booking sites remain, the market itself has become more concentrated. We see improvements in five key areas: The ability to get low fares, viable itineraries, ease of use, customer service and privacy and security policies. Yet we also see many problems remain to be solved. Some are merely growing pains of a relatively new business. Others reveal challenges created by the underlying characteristics of the travel industry itself.

The traditional travel business, operating between the launch of deregulation in 1978 — the airline deregulation era of the late 1970s — and the onset of online travel in the mid-to-late 1990s, evolved with an integrated group of players — airlines, Computer Reservations Systems, travel agents and credit card companies — whose successes were interdependent. In other words, if an airline sold a seat and made money, so did everyone else in the chain.

But the advent of online travel created new business models that altered the relationships among the key players. They became less interdependent and more competitive. Moreover, their two primary goals were now similar: First, generate revenue and build customer loyalty by selling directly to consumers; second, improve profit margins by reducing transaction costs, primarily in marketing and distribution. Instead of sharing customers, now they began to compete for them.

Providers such as airlines and hotel companies sought to reduce reliance on fees to travel agents and Computer Reservations System (CRS) operators by selling directly to consumers through Web sites. In response to this threat to their cash flow, CRSs followed suit, reducing dependence on airline and agent transaction fees (for example, Sabre's creation of Easy Sabre and the Travelocity Web site). The travel agents' response was to build online stores for leisure and business travelers. Credit card companies formed co-branded alliances with hotels and airlines to secure customer loyalty and supplier acceptance, and incorporated travel links into their online payment sites.

As a result, today's online travel market is highly competitive, but also reveals remnants of favoritism among providers and distributors, making consumer education critical. The evolution of the industry has renewed enthusiasm for government regulation. In June 2002, U.S. Transportation Secretary Norman Mineta named a former Maryland state transportation secretary to lead a nine-member congressional commission to investigate the business practices of airline sites and the independent ticket-booking sites to determine their marketplace impact. The commission is scheduled to report back to Congress in November 2002.

Until standards exist — and in the online world, there are few but market forces — consumers who know how the various ticket-booking sites work can score bargains. However, in addition to demanding a certain level of expertise from consumers, airline ticket-booking sites also vary dramatically when it comes to matching bargain-basement fares with viable itineraries, good customer service, and strong privacy and security policies. Flying from New York to Miami via Dallas and back on four different airlines, for instance, is not a viable itinerary.

Each site has advantages and disadvantages. Finding everything on one site can be challenging, but not impossible. It just depends on when and where you're going and when and where you book. Many of these challenges stem from the same issues faced by consumers in the traditional brick-and-mortar marketplace (for instance, one airline's fare getting promoted over another based on its superior position on a travel agent's computer screen). However, online technology can exacerbate these challenges, and can even obscure elements of transactions the consumer could more easily perceive in the "real world."

Furthermore, research by Consumer Reports Travel Letter beginning in 2000 showed disturbing evidence of bias in the way these sites presented fares to the consumer. Research in March and April of 2002 by Consumer Reports Travel Letter and Consumer Reports WebWatch, a project of Consumers Union, supported by grants from The Pew Charitable Trusts, the John S. and James L. Knight Foundation and the Open Society Institute, concludes that while the problems of bias have improved somewhat, issues remain that consumers need to know about. Consumers now have many more choices because the Internet gives more providers more opportunities to create direct consumer relationships and allows a wide variety of pricing. The good news is these choices afford more selection in terms of what, how, when and from whom to buy. But Web technology, coupled with the complex nature of the industry, has created an environment in which evaluating these choices can be mind-boggling.

The independent ticket-booking sites need to address basic disclosure issues — from describing how their technologies work to clearly disclosing business deals they make with airlines that might affect the price of fares, or their position on a screen. The six largest integrated travel web sites confirmed to Consumer Reports Travel Letter that they receive various forms of compensation from airlines, despite the fact most U.S. carriers have recently eliminated the payment of base commissions to travel agencies. Sites also should better separate airline and other advertising from screens of available fares, so the consumer is not manipulated into making a choice based on strategic placement of an ad. Fees should be more clearly disclosed, earlier in transactions — not at the end after a consumer has invested valuable time selecting a flight.

In summary, the Internet travel industry is approaching a crossroads. Until some form of standardization occurs, whether by regulation, market maturity, or both, shopping for the best travel deal online will continue to be confusing at minimum. At the worst extreme, the experience is a little bit like online casino gambling — experienced players can leave the table money ahead, but often, the house wins.

back to top

Introduction

Every second in the United States, $18,500 is spent by resident and international tourists on travel and tourism. The third-largest retail sales industry in America, travel and tourism generated over $580 billion in total expenditures in 2001. (1) The enormity of this figure parallels the scope of changes that have occurred in this industry over the past few years — ever since the Internet expanded from an institutional and educational network to a consumer network, driven primarily by the development of the World Wide Web.

Why did the Internet so dramatically affect the travel industry? What impact do the changes have on travelers? What does online booking mean to anyone who wants to take a trip? To answer these questions, this report looks at the development of the travel industry since airline deregulation in the 1970s, and how the industry developed into one ripe for change when Internet technology arrived some 20 years later. The report explores developments in the 1980s and early 1990s, when computer reservation companies established highly profitable technology networks that supported industry sales and efficient distribution; when thousands of independent travel agents flourished and prospered; when credit card companies enjoyed revenue and profit growth commensurate with increased U.S. airline passenger traffic; and when the number of airlines decreased or consolidated and often earned the lowest margins of anyone in the travel business.

The report then looks at what happened when the Internet offered commercial opportunities for online travel in the mid-1990s and how the race among old and new participants became frantic. Finally, the report discusses what have emerged as today's top online travel companies and what they offer consumers.

1) Source: Travel Industry Association of America

back to top

COMPUTER RESERVATIONS SYSTEM (CRSs) AND TRAVEL TECHNOLOGY

Deregulation meant that airlines that had previously operated under government-set fares which ensured they at least broke even now needed to improve operational efficiency to compete in a free market. While there were many aspects to this, one of the earliest changes was the development of the Airline Reservations System (ARS), its evolution into and proliferation of the Computer Reservations System (CRS), and then into Global Distribution System (GDS).

The history of Airline reservations systems began in the late 1950s when American Airlines began to try to create a system that would allow real-time access to flight details in all of its offices, to integrate and automate its booking and ticketing processes. As a result, Sabre (Semi-Automated Business Research Environment) was developed and launched in 1964. Sabre's key breakthrough was its ability to keep inventory correct in real time, accessible to agents around the world. Prior to this, manual systems required centralized reservation centers, groups of human beings in a room with the physical "cards" that represented inventory (seats on airplanes).

This ability to keep all the data updated eventually led to the ability to price seats on airplanes at many different levels. Initially, however, since there were generally only three reservation classes per flight, sophisticated pricing strategies would have to wait. Other carriers soon followed with their own proprietary Airline Reservations Systems (ARS).|

Quickly, a network concept emerged, which connected the various ARSs together and made them available to travel agents. This became known as the CRS concept. Just as the invention of the ARS enabled the automation of flight and seat control within an airline, the CRS concept automated the reservations process by placing the reservations technology for all airlines on a travel agent's desk, eliminating the need for the travel agent to call the airline to make reservations. This enabled the travel agent to spend more time helping the traveler and enabled the airline to, in essence, outsource the telephone reservation process. This saved the airlines millions of dollars, as the majority of the telephone-reservation work was transferred to the travel agent.

In 1974, an American Airlines executive, Robert Crandall, proposed that airlines jointly create, own and operate a large communication network with global reach for all travel agent offices, and prevent travel agents from taking full control of it. (6) Unwilling to pursue that strategy without antitrust immunity, airlines instead accelerated development of ARSs by expanding them to include other providers and using them as distribution tools for travel agents. In 1976, United Airlines began installing its in-house Apollo CRS in travel agencies; American soon followed. Airlines were under cost pressure, answering calls from travel agents in direct contact with the customer, then inputting reservation information into their own internal systems. Call-center staff performing this work often were unionized. Thus, the airlines' strategy to put the computer and the reservation tools directly on the travel agent's desk outsourced a substantial cost.

What the ARS had done was basically automate the "old model" of a call center at which reservations were written on cards. The technology enabled this function to be distributed not only within an airline, but also to agents and independent businesses completely outside the airlines' control. Soon American's Sabre surpassed United's Apollo in market share and secured a dominant position. The success of the ARS and CRS was clear: In the late 1980s, Apollo's annual pre-tax return on investment had reached 70 percent; Sabre's was more than 100 percent. (7)

Within the next 10 years, European airlines began developing their own CRSs. In 1987, two consortia were formed, and the European-based systems Amadeus and Galileo were designed much like the systems in the United States. Amadeus was based on SystemOne that Texas Air's Frank Lorenzo acquired when he bought Eastern Air Lines. As one industry insider put it, "Frank had to take (money-losing) Eastern in order to get his hands on SystemOne, his real objective in the transaction." Galileo chose United's Apollo system as its strategic partner. In the Asia-Pacific Rim, CRSs primarily operated closely with national airline representatives, with the exception being Abacus, a consortium of Southeast Asian airlines' CRSs.

By the mid-to-late 1990s, the major CRSs essentially became GDSs that travel agents used to check real-time flight schedules, seat availability and pricing information, make bookings and issue tickets. The GDS operators collaborated with a variety of travel service providers such as airlines, cruise operators, hotels, railway companies and car rental companies, in addition to accepting special meal requests, managing seat allocation and performing back-office accounting functions for travel agents. At one point during this period, someone quipped that Sabre might be the most powerful non-military computer in the world.

By the mid-1990s, there were about a dozen major GDSs worldwide. Amadeus had become the world leader after merging with SystemOne, achieving a 27 percent market share; Galileo and Sabre followed, each with 22 percent. After these came Worldspan, formed by Delta, Northwest and TWA, with a 10 percent share, and Abacus and Infini, the dominant CRSs in Asia, with a combined share of 9 percent. (8) (See Figure 4 below.)

back to top

Figure 4

*Market share numbers do not equal 100 % because not all systems were evaluated.

GDS technology developed with four functional components that, while integrated and interdependent, would later serve as points of differentiation when Internet providers entered the market and pulled apart the links of the supply chain. They were: inventory management and display; pricing- and fare-search engines; ticketing and document generators; and database reporting engines.

• Inventory management and display comprised the systems that captured inventory (seats, hotel rooms, cars, etc.) of providers and, through sophisticated algorithms, displayed them on computer screens in response to an agent's keyed-in request. These algorithms were critical because of the physical limitation of the number of flights that could be shown on a CRS screen. Eighty to 90 percent of bookings are made using flights that appear on the first screen; an incredible 70 percent or more of bookings are made using the flight that appears on the first line of the first screen. This phenomenon, called screen position bias, raised regulatory concerns when the owner-operator of the reservation system listed their flights first. Travel agents tended to prefer reservation system owners who provided them with technical support and back-office systems, training and relationship management, causing further concerns. Airlines gave agents incentives to install their proprietary CRSs so they could get "first-line" position and thus generate more sales.
  
  
• Pricing- and fare-search engines were sophisticated systems that would take an itinerary request and, based on a set of rules, determine the fare. The rules were a function of routings, stop-overs, advance purchases, length of stay and a myriad of other factors that, both fixed and variable, were essentially based on supply and demand. The infamous "Saturday night stay" rule was a key marketing tactic discovered by yield managers. It enabled airlines to assign a number of prices to the same itinerary because business travelers, spending company money, were very resistant to spending the weekend away from home.
  
  
• Ticketing and document generators allowed agents to generate a physical or electronic ticket and also queue them to remote locations, such as an airport or out-of-state office, for pick-up. One creative use of this technology was when agents would queue tickets to remote locations for printing where the commission was higher, and then ship them back for actual delivery to customers.
  
  
• Database reporting engines enabled airlines and agents to report transaction activity for a variety of purposes, including financial or accounting uses, trend analysis or passenger searches.

In addition, CRS technology required extensive communications networks to interface in a multitude of technical and geographic environments. Down time on a GDS meant lost revenue for providers, agents and the GDS, as well as frustrated travelers.

In time, more than 80 percent of airline tickets would be sold through CRSs by more than 130,000 travel agencies worldwide. (9) Most of the remaining transactions, such as hotels and rental cars, were also booked through CRSs.

6) Petzinger, Thomas Jr. (1995). "Hard Landing: The Epic Contest for Power and Profits that Plunged the Airlines into Chaos." Times Books: New York.
7) Humphreys, Barry. (1991) The CRSs.
8) Lee, Andrew. Traveling via the Web: The Changing Structure of an Industry. Harvard Business Review Case Study: Center for Asian Business Cases, University of Hong Kong. 1998. Page 4.
9) Flint, Perry. (1998) "End the CRS Oligopoly" in Air Transport World, Vol: 35, Iss: 4 April 1998

back to top

THE MONEY FLOW: ALLOCATION OF COSTS & REVENUES

In the mid-1990s, the economics of travel were fairly straightforward: If a ticket cost $300, the revenue was divided so that the travel agent got about $30, the CRS $10, and the credit card company $6. The airline got the remaining $254, or about 85 percent.

If the same ticket cost $2,000, the CRS fees tended to remain the same, since they were generally segment-based — the example assumes $3.50 per segment for a 2-3 segment trip. Alternatively, the travel agent fee, based on a standard 10 percent commission, jumped to $200, for what many airlines argued was the same or less work if, for example, the higher-priced ticket was for a business traveler, who already knew which airline and flight he or she preferred. Disparities like this motivated the airlines to put pressure on travel agents, seeking changes in the compensation system and reduction of base commission levels.

Revenue Allocation:

• An airline's revenue varied according to the public's demand for travel — highly seasonal, elastic or price-sensitive for leisure travel, and more on-demand for business travel. The airlines' key revenue tool was the yield management system, which allowed them to sell the same seat for 15 to 20 different prices, depending on which market segment the traveler belonged to — business, leisure, price-sensitive, not price-sensitive, etc. They kept low fares from the business traveler by placing a requirement to either stay over Saturday nights or buy the ticket two weeks in advance for a cheaper price. They took what some argued to be a commodity, airline seats or hotel rooms, and priced them differently based on the trip's purpose, day, time, seat location and advance demand in order to maximize revenue. The primary pricing distinction was based on the purpose of the trip.
  
  
• The travel agent's revenue was derived almost exclusively from commissions. In 1995, for example, agents were paid 10 percent of a domestic airline ticket's price, higher amounts for international tickets. To a lesser degree, hotels, transportation and cruise companies compensated travel agents, but these commissions were often difficult to collect as there was no proof the traveler actually stayed where he or she booked. Another important source of revenue, but hidden from the public's view, were supplemental "override" commissions, which airlines paid agents who demonstrated they could move traffic to premium levels, beyond an airline's "fair market share" on a particular route. These agreements are generally thought to have originated after the airlines outsourced reservations activity to travel agents. Doing this gave the agent much more control and "influence" in the traveler's carrier selection process. As override commissions could mean an additional 3 percent to 6 percent, agents were sometimes motivated to influence traveler preference toward the supplier that paid the most. For many agencies, overrides meant the difference between profit and loss, as base commissions were allocated to "covering the cost of operations" and the override "went to the bottom line." Consumer Reports Travel Letter studied the possible impact of these payments in June 2001.

Only about half the travel agencies surveyed immediately provided all the airline and pricing information, when asked, for the lowest-fare non-stop options on 12 different routes. This figure rose to 63 percent after the request was repeated, but 25 percent did not mention all lowest-fare options even after the second request. Furthermore, 12 percent did not provide lowest-fare options at all.

• Charge card companies derived revenue by earning a percentage or discount fee of the face value of any transaction paid for with their card. This fee ranged from 2 percent to 4 percent of the transaction value depending on the merchant's volume of business. In addition to managing the receivable on the part of the supplier, credit and charge card companies promoted the benefits to customers and merchants: Convenience, security, information reporting and increased business.
  
  
• CRSs' revenue came primarily from booking fees charged to airlines, and subscription fees paid by travel agencies and other subscribers to rent CRS terminals and receive technical support from the CRSs. CRSs segmented pricing strategies by region with transaction-based pricing dominating in North America, while net-segment based pricing was used outside North America. In North America, in 1998, the cost-per-net-segment ranged from $3 to $3.50, plus a small fee for cancellations.

back to top

CRS REGULATION

It is important to note that regulators became interested early on by Computer Reservations Systems and possible opportunities for abuse the technology created. Screen position bias, for example, was against free market principles and detrimental to fair competition. Since the CRSs were owned by the airlines and there were no legal restrictions on their administration, market abuses spread. New carriers complained of excessive fees to get their flights listed and established carriers complained of manipulated flight schedules, fare displays and searches. As a result, Congress launched an investigation in 1982, which confirmed the existence of screen position bias and other unfair practices. A comprehensive set of regulations were established in 1984, and then re-issued in 1992. They were designed to address the following four areas:

• Displays — Carrier-specific variables could not be used to rank flight primary displays. Secondary display bias was permitted under certain circumstances. Architectural restrictions, dealing with limiting flight searches, were allowed to continue.
  
  
• Booking Fees — Discrimination in fees charged to participants was prohibited. All participants were entitled to service enhancements.
  
  
• Booking Data — To the degree marketing data was developed, it had to be shared for a fee with all participants. In addition, owner carriers could not discriminate against other systems through "non-participation" in the other systems.
  
  
• Agency Contract Terms — Contractual terms with travel agents could not facilitate unfair competition. Agencies could use multiple systems.

Loopholes in these regulations did exist, particularly with respect to flight information display. Carriers would monitor competitive flight schedules and design their display algorithms so their flights would appear first. For example, if a carrier was being penalized by an elapsed-time algorithm tied to connections over a congested airport, the carrier might simply change the connecting time or flight time, sometimes to wildly unrealistic times in order to improve listing position. (This practice was so prevalent it got its own industry nickname, "time shaving.") Even stranger, carriers sometimes simply invented a new algorithm — say, the number of seats flowing over one connecting airport vs. another — that would serve the same purpose. In addition, CRS rules did not apply to non-U.S. airlines owning CRSs outside the United States. Although the objectives were shared, differences existed between U.S. and European rules. While the United States and Europe disagreed on CRS industry regulation, both maintained shared objectives and open communication. The continued bias in computerized reservation systems was one of two driving forces — the other being fees — that caused providers to seek alternative ways of distributing their products and services.

Within this broad economic framework, relationships between airlines and travel agencies, and airlines and GDSs, were not entirely mutually beneficial. Money was a zero-sum game and as one segment increased its profitability, another tended to suffer financially. From the providers' perspective, particularly airlines, the significant bookings handled by agents came at a high cost. Distribution costs were one of the airlines' top controllable expenses, up there with people, fuel, and aircraft ownership. In 1997, a survey by the International Air Transport Association (IATA) indicated that CRS costs to airlines almost quadrupled in six years, from 2.1 percent of distribution costs in 1990 to 8.1 percent in 1996. And this was at a time when total U.S. distribution costs topped $18 billion. One U.S. carrier claimed its CRS cost increased 35 percent from 1993 to 1998, a period of relatively low inflation and declining technology prices. For the airline industry, distribution represented about 15 percent of the total costs in the late-1980s, and by 1990, was growing faster than passenger revenue. (10)

By the mid-1990s, three important factors had gained enough momentum to drive the aggressive migration to and adoption of Internet-based travel: high distribution costs — and the obvious value-for-dollar question that was raised based on system bias; new technology that offered a cheaper alternative to GDS technology and direct access to customers; and a consumer population receptive and eager to take control of their own destinies.

Just as the airlines outsourced the labor-intensive process of researching and booking travel to travel agencies in the late-1970s, the Internet now provided them an opportunity to let the travelers do the work without the help of any airline employees or intermediaries, thus significantly lowering the airlines' costs once again.
 
10) The McKinsey Quarterly 1996 Number 4, page 180

back to top

Travel on the Web

The travel industry was one of the earliest to go online. Since travel had few geographical boundaries, and, thanks to the widespread adoption of e-tickets, which airlines aggressively pushed, the airlines faced none of the logistical issues of online product retailers such as shipping and variable tax-collection schemes, the travel industry was uniquely suited for the Web. As this report seeks to demonstrate, the lack of rules presented a major problem. There was no governing or marketing body to regulate rates, or online travel services that enabled consumers to efficiently navigate all the alternatives they had.

Bill Gates reportedly once quipped that Microsoft started Expedia because no other industry was as complex as travel, with so much constantly changing electronic information and consumers who wanted to become personally involved in the reservation-booking process.

With existing players and new entrants trying to capture a slice of the online market, new travel-related Web sites were springing up or reinventing themselves constantly. Many airlines, hotels, car rental companies, CRSs and national and municipal tourist organizations went online. Participants attempted to capitalize on the opportunity in travel by developing products aimed at attracting and retaining customers. To capture traveler loyalty, travel agents and CRSs built "consumer-friendly" front-end systems for existing information systems. They also leveraged relationships with emerging online service providers to work with business and leisure travelers in a cost-effective manner. In terms of size, complexity and sophistication, these first-generation efforts, by such players as EasySabre, Prodigy and ITN (reviewed in more detail later) are far removed from today's mammoth sites.

Airlines reduced the need for intermediaries by offering direct-access software and encouraging business and leisure customers to purchase tickets directly from airlines. Each player in the travel supply chain would have its position challenged. Ability to adapt quickly would determine survival. In this new world, both the economics and even the players would change:

• Airlines reduced, and in some cases eliminated, costly intermediaries such as travel agents from their distribution chain. In 2001, there were 15 percent fewer travel agents in the United States than five years earlier. Airlines re-structured their distribution compensation agreements by segmenting them. Traditional transactions got one rate, electronic another. These categories were further segmented based on who, how and what was booked. In today's environment, travel agents provide less influence on carrier selection than they did in the past. Since the airlines no longer thought agents steered consumers to airlines, the airlines adjusted their compensation to virtually eliminate base commission payments in lieu of "pay-for-performance" structures. These pay-for-performance remnants of the "override" world continue to put financial pressure on travel agents' loyalties.
  
  
• Travel agents have redefined the way they charge consumers, in many cases unbundling their services. This may include charging a $25 service fee for issuing a ticket or a nominal fee for changing it. In the corporate travel arena, travel arrangers might pay for information management, on-site passport or back-office processing. Many travel agents invested in electronic servicing capabilities either independently or with technology partners, which had the two-fold objective of reducing their own service costs as well as providing entry into the new electronic market.
  
  
• Global Distribution System (GDSs) continued to consolidate and diversify their operations by unbundling their services and expanding their product offerings into other transaction-processing and information management services. These were meant to serve customers other than airlines, who were trying to reduce or eliminate GDS fees all together.

back to top

THE ONLINE TRAVEL MARKET

So how is the online travel market defined? In 1997, it was estimated to have made up 1 percent of the total U.S. travel market. Today it has grown to 11 percent, valued at over $20 billion. (11) The market was already growing prior to the terrorist attacks on New York's World Trade Center and the Pentagon in Washington, D.C., on September 11, 2001, and since then consumers have continued to rely on these sites both as "looking" and as "booking" tools.

The demographic profiles of those who made online plans at the emergence of Internet travel in 1996 were more likely to be employed, educated to the postgraduate level and a professional or manager. Five years later, online travel has gained mainstream popularity, and there are thousands of sites that offer travel information and services. They range from direct providers (e.g. airlines, hotels, ground transportation) to support services (e.g. tour organizers, travel and trade publications, hotel management companies), tourism development organizations, eco-tourism coordinators and many more. Depending on the publisher, travel Web sites could be categorized as service providers, destination related or Internet travel agencies (sometimes virtual branches of traditional brick and mortar agencies such as American Express, or virtual branches of Internet portals like AOL and Yahoo!).

• Online service providers are branded sites developed and operated by airlines, hotels and others that want to sell directly to consumers. These sites are similar to auto dealerships in that they focus on selling the products of a particular supplier — American Airlines, for example, will not direct a consumer to United for a lower fare or more convenient service. That means if you use a supplier's site and are concerned about price, you might have to shop around to get the best price among other supplier sites. For example, based on actual research, on one routing an airline site quoted the lowest fare as more than $1,200. A check on a travel agent's site for the same trip quoted a $400 fare on the same carrier, with a connection via a city along the non-stop flight path that could not even be found on the carrier's own site.
  
  In early days, provider sites might only have offered information. Today, most offer a comprehensive array of products and services online. Major airline sites offer customers reservations, electronic tickets (e-tickets), seat selection, in-flight merchandise, reward points and sometimes discounted fares unavailable elsewhere. In addition, they may offer lodging, transportation-package deals and cruises through their alliance partners. For example, American Airlines offers and redeems AA miles when a member makes a purchase on AOL. Branded airline sites are currently the fastest-growing segment of online travel providers, up 26 percent in February 2001. The most popular site (as measured by most unique visitors) is Southwest Airlines. One reason Southwest is so popular is that it doesn't issue tickets, and historically travel agents could not book them via a CRS because Southwest refused to place its information on the systems. Today, Southwest's inventory is available online only at its own site. United and American are the largest airline sites in terms of sales.
  
  
• Destination sites provide information and services about a country, city or area, including details on transportation, accommodations, sightseeing, dining and local cultural events. They may or may not offer advertising, online booking capabilities or discounts and coupons. Since destination marketing is usually the responsibility of the national or local tourism boards, their Web sites usually embody this tone. Their goal is to generate visits to the area, so they are less concerned with the mechanics of how the traveler chooses to get there.
  
  
• Internet travel agencies are Web sites that expand traditional agents' offerings. In addition to selling regular travel services such as air tickets and hotel rooms, they also offer travel tips, destination information and other services. Many large traditional agencies such as American Express and Liberty Travel have extended their shops to the virtual marketplace. Portal travel sites such as AOL and Yahoo! link customers looking for travel services to sites that may appear on their portals, but generally are powered by one of the big Internet travel agencies listed below. The last category for Internet travel agencies is bidding sites, such as Priceline. There, deep discount travel is available, but the travel provider's brand (usually an airline) is hidden until the purchase is complete.

Consumer Reports Travel Letter in 2002 evaluated six of the largest and most well-known sites: CheapTickets, Expedia, OneTravel, Orbitz, TravelNow, and Travelocity. (See Figure 8 below for a Summary Table describing the sites.)

back to top

Figure 8 - Major Virtual Travel Agencies - 2002

For consumers, online travel and its new world of self-service has brought convenience, access, speed and control. It has also brought confusion and lack of consistency — albeit a consistency once confining, but at least one that let a consumer identify a good deal or not. Now, customers whose experience with booking travel was traditionally insulated from the complexities that lay below the surface are now given more choices about how and where they book and more services from which to choose, and yes, pay for. One key theme in this latest evolution in travel has been a redefinition of who the customer is at each stage of the process, and aligning costs and revenues accordingly. In the old world, the customer paid nothing explicit for travel services. The cost of the service was bundled into the cost of the air ticket. In today's environment, where the airlines have moved to zero commissions for travel agents, the cost of services — in the form of service fees — is now apparent. The airlines contend the consumer always paid for these services, but the money is no longer driving the airlines' financial statements.

Hence, leisure and business travel consumers have much more choice today in what and how they buy. It's not just selecting the destination, date, time, carrier and seat, but whether they want to research a trip independently or enlist an agent to help, receive an electronic or paper ticket, self-select the destination and travel logistics or have them selected for him based on excess inventory at deeply discounted prices. Today, travelers can tour hotel rooms through their Web browsers, find online references from other travelers and comparison shop in the most efficient way ever.

But given what we have discussed so far about how the air travel industry evolved in the last half of the 20th century, of the airlines' dominant role in all phases of its growth, and of how regulators in the 1980s felt compelled to establish rules where none before existed, what do consumers need to know about this new, fast-growing frontier in which the power appears to be in the hands of the people?

Let's look at this new world of Internet travel in more detail.
 
11) Source: Jupiter Media Metrix; http://www.jmm.com/xp/jmm/press/2002

back to top

Consumer Experience

Consumer Reports Travel Letter last evaluated four large independent sites-Cheap Tickets, Expedia, Lowestfare and Travelocity-in October 2000. Though the findings were not conclusive, they did reveal some evidence travel sites may not always be totally objective:

• On Travelocity, advertised airlines dominated flight listings.
  
  
• On Lowestfare, many TWA flights with inconvenient itineraries (obtained through a contract fare deal) were repeatedly listed first.
  
  
• On all four sites, certain airlines with viable itineraries for routes tested were not listed at all.

As previously discussed in this report, the concern about bias dates back to the 1980s, when the U.S. Department of Transportation (DOT) devised rules for Computer Reservations Systems after receiving complaints about biased displays from travel agencies and airlines. The DOT was clear about regulating only airline-owned CRSs, so that such systems couldn't unfairly aid their sister-company airlines.

None of the travel Web sites in this series of tests were owned by an airline, although all four accepted advertising. And in some cases, airlines pay for more prominent placement. Travelocity presents "featured airlines," which receive full-color advertisements linked to specific cities or airports. When you request a list of fares, Travelocity asks if you would like flights offered by the featured airline, or choices from other airlines, as well.

In Consumer Reports Travel Letter's 2000 testing, the featured airline on Travelocity was listed first 48 percent of the time and dominated other listings. In nine separate tests, each recording the top nine flight choices, the featured airline flew at least one leg of every trip that Travelocity posted. Many of these trips involved convoluted itineraries melded together with at least one other airline.

Lowestfare's contract agreement with TWA seems to have influenced its flight listings: In tests, that airline was listed as the first choice 50 percent of the time (in 27 of 54 tests), when no other site listed TWA first more than 23 percent of the time. The TWA routings sometimes involved connecting flights when other Web sites offered non-stops.

back to top

MISSING AIRLINES

Another concern was raised when cross-referencing test results revealed that certain airlines listed in the Apollo Galileo benchmark as offering the lowest fare and a viable itinerary were not listed on some Web sites at all. Spiri