WHY FRONTIER AIRLINES FAILED An Analysis by Dena Andrews, et al HISTORY Frontier Airlines was officially born on June 1, 1950 resulting from the merger of Monarch Air Lines, Challenger Airlines and Arizona Airways. The new company’s home base was in Denver, Colorado servicing the Pacific Southwest Region. During the 1950’s new routes were added to their distribution network, as well as new 44-passenger Convair 340 aircraft to serve most of Frontier’s system. In the 60’s the company introduced a special fare plan called “21” to cut the cost of travel, resulting in a 26% percent increase in passengers enplaned. All time records were experienced in 1963 with passenger boarding up 44% and growth exceeding all other 23 regional airlines in the United States. Frontier later became the technical leader by acquiring jet power aircraft including the Boeing 727 and 737, along with being the first to introduce a computerized reservation system. To further augment their growth, Frontier acquired Central Airlines advancing their routing network to 114 cities covering the Midwest, West and Southwest. The 70’s brought a slowdown in the industry resulting from depressing economic conditions and excessive competition. However, Frontier Airlines was able to continue expansion of their routes, including the addition of international destinations with Canada and Mexico. The airline’s superior customer service lead to them outperforming their competitors and achieving an all time high in paying passengers. This period also introduced the Airline Deregulation Act of 1978, providing an environment full of opportunities for stronger competition among the carriers. In 1980, Frontier Airlines became the prime subsidiary of the newly created Frontier Holdings Inc., and became an all jet airline with the retirement of the Convair 580s, which were replaced with McDonnell- Douglas MD-80. But the 80’s also marked a decline in traffic with the termination of service to 11 smaller cities, fierce competition and financial distress. Net losses of $13.8 and $31.3 million were incurred in 1983 and 1984 respectively. The prevailing environment presented by deregulation and technological innovation imposed a managerial challenge for Frontier, as they sought alternatives to improve operations and successfully compete in the industry. SIGNIFICANT CHANGES In 1978, the Airline Deregulation Act brought about a dramatic change in the industry such that airlines could expand their routes simply by giving a ninety-day notice. This environment encouraged fierce competition among some carriers fighting to dominate the Denver hub. Between this period and 1982, Frontier dropped 39 routes from their network and added 29 new cities with further emphasis placed on their Denver hub. Given the changing environment, Frontier placed additional attention on restructuring and gaining a competitive edge needed in a deregulated market. Mr. Lowell Shirley, who had worked with Frontier’s data processing system in the past, was recruited in 1981 to help Frontier achieve its objectives. In 1983, Lowell Shirley restructured management’s goals and drew up four commitments: a) Reduce maintenance of old applications from five person-years to three b) Reduce the cost of Frontier’s voice communications network by $30,000 per month c) Relocate data entry to user departments d) Relocate hardware to the new facility and complete a new set of applications HARDWARE Frontier’s hardware included three IBM System 38s, one IBM Series 1, and several stand-alone minicomputers for marketing, budgeting, maintaining personnel records and domiciling flight attendants. The System 38s were employed in production, software development and remote job entry, while the Series 1 handled remote computing and communications routing using X.25 packet-switching protocol for linking outlying stations. The advantage of Systems 38s was that they were modular and could be expanded for growth. Control Data Corporation maintained the hardware that facilitated Frontier’s reservation system until 1976, when a decision was made to begin purchasing reservation services from Continental Airlines. The use of standardized IBM microcomputers at the managerial level was also an important topic introduced under Frontier’s policy. However, only divisional data was allowed to be used in each microcomputer, and all other required applications needed to be approved by technical specialists. SOFTWARE Technical personnel, most of whom did not have airline experience, developed distributed system software in-house. Some of the applications were automated but not linked, such as general ledger and accounts payable. These systems focused on compatibility rather than integration, and only basic spreadsheet analysis packages were commonly used. Users who wanted new applications had to follow the hierarchical organizational structure to obtain additional equipment and personnel. RESERVATION SYSTEM Frontier’s reservation system, SENTRY, was run on a System 38, which housed over 70,000 fares in its database. It was later enhanced to include advance seat selection, assignment and boarding pass production. To compliment its reservation service, Frontier became a part of American Airline’s frequent-flyer program. Frontier tried to market its reservation system to travel agencies but was unsuccessful in competing with large carriers for travel agent acceptance. The company then decided to become a co-host under United Airline’s APOLLO System, a widely accepted system in the industry. Their initial bid was rejected on United’s claim that APOLLO could not handle the additional carrier; although three other airlines became co-hosts during that period. Finally, in 1982, Frontier became a co-host under APOLLO that resulted in significant problems for the company. COMPETITORS’ INFORMATION SYSTEM After the deregulation, the use of computer technology gained a great level of acceptation among travel agents, considering it made easier their task of obtaining schedule and fare information. However, the use of carrier-controlled computer reservation systems had anti-competitive effects on smaller airlines. Some systems were designed to promote the flights of the carrier that sponsored the reservation system before the competing airline’s flights. The following exhibit provides an example of the effects mentioned above: If a passenger requested an early morning flight between Detroit and Denver, the travel agent using Sabre® would have arranged Frontier’s flight No. 403 for the customer. However, the APOLLO system would have favored United’s flights. According to the report submitted by Frontier Airlines to Congress during the antitrust investigation, 148 out of 198 (75%) travel agents in the Denver area were subscribed to United’s APOLLO, a significant factor that strengthen Frontier’s financial distress. PROBLEM ARISING IN 1983 The first problem arose from Mr. Shirley’s decision not to run a reservation system. Because of this, Frontier had to rely on the systems of other airlines. United Airline’s APOLLO reservations system was the most widely used system. However, United refused to allow Frontier to become an APOLLO co-host for competitive reasons and Frontier was left without a reservations system that linked it to travel agents. American Airline’s Sabre® system was the second most popular reservations system, but United was discouraging its use to travel agents, claiming that it was an ineffective way to do business. Not only was the costly battle to gain entry to APOLLO an economic setback for Frontier, but it also contributed to Frontier’s backsliding in technological advancement. Upon entry to APOLLO, Frontier discovered that it lacked access to pertinent real-time information that United Airlines would not release to co-hosts, and the information that was shared to co-hosts was a month old. This gave United Airlines an unfair competitive advantage. Further, travel agencies were being given erroneous information that Frontier flights were booked when they were not, a situation that contributed to a decline in Frontier passenger bookings. Another problem resulted from Shirley’s analysis of Frontier’s problem. Shirley concluded that only 5% of data came from outside sources without prospecting what the new data requirements of a deregulated industry were. His analysis was also flawed by his lack of knowledge of the changes that occurred in the industry since his earlier departure. The resulting solution was a distributed system network that lacked the integration capabilities required to increase the company’s efficiency and agility in the marketplace. ALTERNATIVE SOLUTIONS The following strategies can be applied pending the decision of The Department of Justice and Civil Aeronautics Board regarding United Airlines antitrust practices. These solutions focus on providing alternatives using the technology widely available today. Over the period 1978-1982 operating expenses were increasing significantly while revenue increased only moderately. Additionally, subsidy for unprofitable routes was inconsistent and with deregulation would be eliminated. Therefore, Frontier needed a system that would allow them to closely monitor their cost and operational profitability. Using decision support system, Frontier can closely monitor internal financial performance of their routing network. Decision support tools provide invaluable scenario analysis in the absence of external information. These tools interrogate databases using what-if analysis. Such questions as what if competitors decrease prices by 5%, 10%, etc. and what if subsidy is no longer provided on particular routes could be explored. Managers could project what dollar of cost containment must occur given the various predictions. In the long-term, such information would provide assistance to executives about the overall financial situation of Frontier, covering the existing gap between general ledger and accounts payable. Also, their reports would indicate what effects are being experienced in the internal environment and critique the reasons. An excellent choice is the American Analytical Information Management System (AAIMS). American Airlines and other industry partici pants currently use this system. AAIMS supports a variety of airline decisions by analyzing data collected on airline aircraft utilization, seating capacity and utilization and traffic statistics. For example, it produces forecasts of airline market share, revenues, and profitability and helps airline management make decisions on aircraft assignments, route requested, ticket classification pricing, etc. Furthermore, this system would reduce redundant work, resulting in improved employees’ performance and better decision-making. LONG-TERM SOLUTION Regardless of the outcome of the antitrust investigation, Frontier needs to adopt a strategy that would provide a long term competitive advantage such as providing superior customer services and better travel agents relationship. VPN (Virtual Private networks) could have given accomplished Frontier a real advantage over United Airlines and all of the rest of its competitors. VPN would have provided access to real-time data by all employees of Frontier, and data sharing would have been not only possible but also sensible. This also would have eliminated the reservations system portions with a travel agent or directly with Frontier. “ Thus, anyone with Internet access worldwide can participate in this adaptable WAN (Wide area network). With it, you can provision the network on-the-fly”. From an interview with a computing technology specialist, this solution would not be practical. VPN would provide travel agents access but require them to log on to a different system. It is not convenient and is time consuming thus decreases VPN’s attractiveness. A strategic alliance with American Airlines to use their Sabre® reservation system would provide Frontier an effective way to gain higher visibility in the market. The system is run on IBM’s 390 an air-cooled parallel mainframe which operates half of its airplane ticket pricing work load and is 100% reliable. This system was the first to install Parallel Enterprise Servers (PES), which reduced transaction-processing cost and avoided application conversions and data center investments. The parallel systems worked without problems during a fare war in 1994 when Sabre®’s message rate was at an all time high. Using online transaction processing, the system generates real-time data that helps users to provide superior service to their customers and travel agencies. OLTP would add value to Frontier’s services and thus would give them an important tactic to differentiate themselves from their competitors. Sabre® software handles scheduling, yield management and pricing and its technology has been cutting edge. CONCLUSIONS We believe that our proposed strategies would have placed Frontier in a better competitive position in the deregulated market. Frontier relied solely on the outcomes of the Civil Aeronautics Board ruling on United Airlines’ antitrust violations and did not search for alternative solutions. They failed to address increasing operating expenses, which resulted from not identifying their internal problems caused by Shirley’s wrong decision-making. Further, United’s APOLLO system was not the only option for Frontier. Our research showed that Sabre® was not biased in displaying other airlines’ schedules. Co-hosting the system with American Airlines, Frontier could also pursue membership under the AAdvantage Frequent Flyer Program. Advantage provides Frontier an opportunity to build customer loyalty by creating a cost of switching to other airlines. (The preceeding article was edited and abridged for clarity and space. The following background information was taken from the Frontier Online Club message board.) Hello Mr. Lamkins, I am a graduate student at Stetson University involved with a group of students doing an analysis of the failure of Frontier Airlines in 1986, from an information technology standpoint. Our case alludes to the fact that the airline failure may have resulted from Frontier's inability to compete with the United's superior reservation system - Apollo which was widely accepted among travel agencies. Would you be kind enough to provide your feed back on this position and recommend any others sources that may shed some light on the situation from an information technology standpoint. Thank you so much. -Dena Andrews Hello Dena, I worked for Frontier for 22 years: 1964-1986. I did everything but "fix'em or fly'em". Frontier went bankrupt for a variety of reasons. I don't believe the reservation systems in place at the time were significant. The main reason Frontier went bankrupt is because three airlines had hub operations in Denver and the market would not support that many. When Continental Airlines filed bankruptcy in about 1984 we at Frontier felt assured of our future. Unfortunately, Frank Lorenzo manipulated and misused the bankruptcy laws to continue operating Continental at a cost that no one could match. United had other markets to make up for Denver; Frontier did not and we bled to death in a few years in spite of massive attempts by the employees to save it. -Jake Lamkins Hi Dena, My name is Rudy Castillo. I started my airline career at Frontier Airlines in 1983 as an airport agent and then moved to reservations. I was really into the res system, so when Frontier died I went to work at United's Apollo travel agency help desk. I never much cared for United, so after two years I moved on to American Airlines and Sabre. I worked in many different areas within Sabre for 12 years. I just moved to National Airlines in Las Vegas, who is hosted in Sabre. I am the Manager of IS Vendor Relations, what that really means is that I am the Sabre guy at NAL and I manage our contract with them and the other GDS'. Now, I will attempt to give you a little information on what automation was like at the original Frontier Airlines. Frontier was a CCS (Continental Computer Systems) user, the system is now known as Shares. CCS/Shares was designed for basic airline use and didn't have all of the bells and whistles that Sabre and Apollo had/have. It is important to note that Sabre and Apollo have a lot of really cool functions because they are marketed to travel agency subscribers. Agencies generate a huge amount of revenue from booking fees, so any investment to add enhancements and new functionality virtually pays for itself. CCS/Shares really did not do everything Frontier needed. Other airline schedules were sometimes hard to find or were not in there at all and it's itinerary pricing was very limited. Frontier flew into Canada and Mexico and to my knowledge the system never calculated the taxes correctly. This was never really an issue because Frontier's reservations agents were VERY proficient in fares and taxes. Sabre and Apollo travel agency subscribers had "direct access" to Frontier. They also had the capability of requesting seat assignments and generating boarding passes. "Direct Access" in those days was nothing like what we have today. It was a one way link and it failed very frequently. In 1984 "direct access" was cutting edge technology in the airline industry. The product we have today is truly awesome! (I can provide more info on this if you need/want it.) Shares is alive and well working for quite a few carriers. It's largest customers are Continental and America West. I honestly don't think automation was even a factor in Frontier's demise. The main reason was the fact that Denver simply could not support three hubs. -Rudy Castillo Hello Rudy, Thank you so much. Your input is invaluable. The Harvard Business School case study on Frontier Airlines totally focuses on the information technology side of things. It talks about the Apollo system displaying other airline schedules inferior to United's, and the pressure United placed on the travel agencies to book most of the reservations to their carriers. Additionally, it outlines Lowell Shirley's decision not to integrate the systems. From this standpoint my group has to make a IT recommendation for Frontier Airlines. We are challenged to assume that this problem arose in 1983 but the technology of today can be offered as possible solutions. Frontier's Sentry system did not provide realtime data which perhaps would have aided in better managerial decision making regarding pricing changes. We are operating on the belief that when bookings are low days/day before departure some airlines tend to drop fares to try and increase their passenger load. From this perspective we are formulating our recommendation. -Dena Hi Dena, Your message gives me a better idea of what you are looking for. In the early days of what is now known as a CRS/GDS, airline terminals were installed in travel agencies to generate more revenue to the host/owner airlines. The schedule/availability displays were VERY biased with Apollo favoring United and Sabre, American. At first only the top producing agencies for each carrier were considered for automation. As an incentive to the agency, the host carrier offered commission over-rides based on market share. This did put carriers like Frontier at a major disadvantage. These over-rides represented big bucks to a travel agency! The travel agency business was very different in those days. It really was easier to book the host carrier, and the distribution cost to the airline, only 10% of the ticket, was definitely cheaper. This eliminated the costs associated with a res agent taking the booking and an airport agent issuing the ticket. Revenue management in those days was poor at best. The tools that exist today, such as MIDT and BIDT which are provided by GDS, were only a dream. Off-line historical data was only as good as a carriers record keeping. Yield management was usually based on an individuals "best judgment" rather than real hard data. With what we have today, we can clearly see that load factor and the actual yield have nothing to do with each other. The spin off of the GDS from the owner airlines opened up a whole new world of opportunity for both the GDS and the worlds airlines. What we have today is the closest we have ever had to a level playing field as far as automation is concerned. If Frontier had survived the Denver hub war, it's biggest challenge would have been to modernize an aging fleet and keep up with the automation demands of the industry in the late 80's and the 90's. Frontier's strongest asset was it's employees, there is nothing they wouldn't have done to keep their airline alive! To this day, most are very emotional about what happened so many years ago. -Rudy The replies to the graduate student have been interesting to read and there appears to be a common belief that one of the significant factors behind the demise of Frontier was the situation of three airlines having their hub in Denver. This brings up a "what if" thought that has often run through my mind. I remember once reading the story behind Western Airlines moving their hub from Denver to Salt Lake City and the behind the scenes events surrounding this move that Western Airlines made. Something that I remember reading in this story was how Salt Lake City wanted to have a major hub located there and how they first approached Frontier Airlines about relocating their hub from Denver to Salt Lake City. When Frontier said no to this proposal, Western Airlines was approached next and they made the decision to make this move. This move ended up working out well for Western Airlines and their hub in Salt Lake City was successful up to the time they were taken over by Dela Airlines and now this continues to be a major hub for Delta. My "what if" thoughts have to do with what may have happened to Frontier if they had accepted this offer by Salt Lake City. Would their future been much better if they had made this move? Would they still be around today if they had made this move? It would be interesting to see the thoughts on others in this group on this! -Bill Buse I don't recall much about the SLC hub idea that you mention, Bill, but I do recall many times hearing Ryland talk about his "dumb-bell hub" concept. Usually it meant making MCI a second FL hub but sometimes he mentioned other possibilities - always pairing it with DEN. Nothing ever came of it, of course. I don't think there was ever serious consideration of quitting the DEN hub. -Jake Lamkins Hi Dena, I'm the editor now of a quarterly newsletter we put out to the employees, family and friends of the "old" Frontier Airlines. I helped you out several months ago on a report you were working on regarding why Frontier failed. With your permission, I'd like to run your report in the newsletter - with appropriate credit of course. We just ran a long two issue article by an ex-employee about what caused the airlines' demise. I'd be interested in showing how it looked from an objective outsider - you. If it's okay, please let me know along with a brief bio that I can include in the credit. I'd be glad to send you a copy of the issue when it runs too. -Jake Hi Jake, of course you may use the report. It was compiled by Wladimir Simosa, Kathy Torrunguang and myself, Dena Andrews. We were graduate students at Stetson University, Deland, Fl. The work was completed for our Computing in Management coursework. The research was insightful and we were thrilled to receive responses from individuals like yourself that were directly involved with the airline. Thanks. -Dena