In the Enterprise search market, which was expected to grow, there was still fierce competition. The amount of information is increasing in all companies, and the need for information sharing is of growing importance. There are several companies with different technologies and no one knows who will win this battle. In this phase the companies are forced to choose a path and many challenges lies ahead. FAST Search & Transfer is competing in this industry; this case is written to highlight the challenges and opportunities for FAST Search & Transfer in this rather new and interesting industry.

1.0 History and development of FAST Search & Transfer ASA[1] 

FAST search and Transfer is the second largest IT-company at Oslo Stock Exchange with a market value of approximately 1.9 bill NOK or USD 265 million (by April 2003). The company has about 160 employees and have activities in Norway, Italy, France, Germany, UK, the US and Japan. Revenues in 2002 were about USD 46 million. In brief their business is solutions for digital information search and digital information sharing.

The history of FAST has its foundation from research at the Norwegian School of Technology and Science (NTNU) in the beginning of the 1990’s. This research focused on the fundamental problem of searching and filtering large volumes of data by many users. People like John Lervik, Espen Brodin and Tor Egge initiated and developed this research; some of the researchers were at a PhD level while others worked on their master’s thesis. Over several years of research the technologists at NTNU built up a world leading expertise on this field.

At the same time a technology interested and entrepreneurial archeologist, Hans Gude Gudesen, wanted to make a software program which could make it possible to search for registered archeological items. He got in contact with John Lervik and Espen Brodin, currently working on such a search engine. Gudesen had three years prior to this started a company called Opticom, which made some research on the idea that information stored on plastic could store much more information than the currently magnetic technology. Through this company Gudesen had got in contact with Thomas Fussell and Robert Keith, two former investments bankers from London. These three people saw the commercial potential of the search technology from NTNU. They invested in the technology and started up Fast Search & Transfer as a subsidiary of Opticom in June ‘97, together with the researchers John Lervik and Espen Brodin.

FAST Search & Transfer started up carefully with only three employees working with development and improvements of the technology. It was listed on the OTC-exchange, the gray market in Norway, where Opticom was the majority shareholder with 56% of the shares. The employees and a few investors kept the rest of the stocks. By the end of 1998 FAST got a breakthrough when they signed an agreement with the worldwide portal Lycos, an agreement where Lycos would use FAST’s search engine at their website. Lycos was one of the largest players on the Internet at that time. This agreement, together with an agreement with Dell, gave FAST increased attention in the market and fueled FAST financially. During 1999 the company really became commercialized. FAST experienced a tremendous growth, and by the end of ‘99 the number of employees had increased to about 60. At the same time the stock value had increased from 40 million NOK (‘98) to about 13 bill NOK in the end of 1999. Still, however, most of their focus was on development and the new employees were mainly engineers working on further development of their technology. The agreements with Lycos and Dell gave FAST prestige, but the revenues were still very low.

FAST had experienced an increase of value of unbelievable 6066% in 1999[2]. By 2000 the world was at the top of the Internet bubble, and FAST had despite quite low sales no problems with funding. It was important for FAST to develop quickly and expand their business. During 2000 FAST expanded globally, something which involved a tremendous expansion of the organization. FAST build organizations in 5 countries; UK, the US, Germany and Japan besides Norway, which still housed the R&D department. By the end of 2000 FAST had grown to almost 300 employees. This tremendous growth had led to a huge growth in operating costs, but increase in sales didn’t match the increase in costs resulting in very high losses.

In the beginning of 2001 FAST reduced their focus on organizational development and R&D, to focus more on sales and marketing. A company of this size could not live on the investors’ mercy forever. The crazy Internet gallop of 99-00 had led FAST into a huge growth and they had focus on a broad range of business areas, areas previously and partly started to satisfy inpatient investors afraid of missing the opportunity of the Internet revolution. This strategy had led to an increase of products, many of these outside their original business model. FAST had problems to concentrate on where the money actually was. Understanding this, the management developed a more clear business model into areas where it was possible to earn money. They removed their focus on MP3 search (profitable, but legally problematic), WAP search and Multimedia search and started to focus more on their enterprise business. After these adjustments FAST had two main business areas, Web-search and Data search for enterprises. These adjustments in combination with more attractive products as well as a more fit organization led to an increase in sales of 634% in 2001. 2001 was also the year when FAST finally got listed on the Oslo Stock Exchange main listing after some delays. Their plans also included to get listed on NASDAQ, but the listing was cancelled due to the economic downturn and the negative attitude among investors towards technology companies. This delay and the cancellation from NASDAQ gave FAST, to some extent, a bad reputation among investors as well as analysts, resulting in negative impact for FAST in the stock market. It was now even more crucial to make money. In the last quarter of 2001 FAST downsized with 60 employees in an attempt to reduce costs. This move together with actions led to a reduction of about 30% in costs in 2001. Still, however, FAST was unprofitable. The bad results, the negative attitude in the stock market and the adjustments of stock prices in most of IT/Internet companies led to a huge decrease in the stock price of FAST. This led to some changes in management as well as the mentioned adjustments in their business areas. In this period the founder and former CIO John Lervik took over as CEO.

Year 2002 was a milestone for FAST. It was their first year with profits. This came as a result of increased sales in combination with a lowering of costs of about 25%. It was also positive for FAST to experience that search technology did not face the same decline as other parts of the IT sector after the IT-collapse. Instead the level of data in huge corporations and the level of information on the Internet have continued to grow. FAST started to focus even more on data-search solutions for big corporations in 2002. Big corporations have huge amount of data and are willing to pay for solutions, which can help employees and customers to categorize and find the right information at a little time as possible. Today search engines are the best available technology to do this. Several big corporations like France Telecom, Reuters, IBM, eBay and Telenor found FAST solutions to be very helpful. FAST also increased their market share in 2002, to be around 20% of the industry’s total revenues, up from about 12% in 2001 and 2,5 % in 2000. (Market share among their main competitors; Verity, Autonomy, Convera) (Exhibit 2,1)  

In February 2003 a major phase found place when FAST sold their web-search technology, Partnersite and their webpage alltheweb.com to the American company Overture for $70 mill, with additional incentive payments of $30 mill over 3 years. In this transaction 50 of the employees in the web-search part followed the acquisition and started working for Overture. Web search had previously gained about 25% of FAST’s total revenues and had never been profitable, thus making the agreement popular among investors and analysts[3].

From this point FAST is putting all their energy in Enterprise-search through their Data search solutions. This strategic choice gives them many challenges in an area where they are facing more mature competitors with longer experience.

On the positive side, they have received a delicate problem of what to do with the positive cash flow from the sales and profits, which is about $100 mill. They have got different choices and prospects, ranging from acquisition of new companies, increase their R&D, give a high dividend payout, go into new areas of business, or build a larger sales and marketing organization etc. Anyhow, FAST’s business model has changed and they have a lot of strategic challenges ahead in their fight for customers. Knowledge about their technology, customers, competitors and organization is crucial to understand the industry and the market opportunities. 

2.0 FAST’s Technology. 

FAST is a technology driven company, once started up with an idea of a world leading technology. Researchers at the Norwegian School of Technology and Science (NTNU) developed FAST’s core technology. The researchers at NTNU were early developers of their solutions. In 1997 Tor Egge (student at NTNU, now one of the researchers at FAST) started a FTP-search website which became the most visited Norwegian website with about 16 mill hits that year. This site was developed as an extension of a thesis written by two students, Hugo Einar Gunnarsen and Stig Sæther Bakken in ‘94. In ‘97 Tor Egge was operating this FTP-search engine from his own dorm at NTNU, among some enthusiasts it was rated as the best web-search site in the world[4]. The research-environment at NTNU was one of the world leading environments for search engines at this time. Today FAST have developed this technology further, and also patented parts of it to protect their technology from replicas.

The core of FAST’s technology is their linear scaleable platform, which makes it easy to combine several low cost servers into a network able to deal with whatever demand for performance there might be (see exhibit 5 and 6). Linear scalability of capacity means that you will increase the capacity linearly by the total amount of servers plugged into the network. Larger systems might combine several search servers with a number of separate servers for data aggregation, content pre-processing, and front-end query analysis and result set processing (the final result set given to the user) to handle heavy query volumes. The search performance is possible to measure with the amount of information possible to search for, how often the information is updated and how many users there are.[5]

However, the crucial point is to know what the FAST technology does for the customers. When users ask a query at a search engine, very often the users aren’t able to specify well enough what they are looking for. This means that the search engine has to deal with uncompleted searches, and on the other hand help the users to specify their query in a better way. FAST is dealing with this problem by using an advanced analysis of the queries done by the users. In the analysis of searched information and linguistics, the software is narrowing the search by using synonyms, misspellings, detection of phrases and names crucial to the application, and finally a detection of grammar mistakes.[6] There are four key processes that work in tandem to assure the highest level of relevancy of search results[7].

1. Understand the content; The system identifies names, reduces words to their base forms, extracts content from pages and categorizes each document based on its subject matter.

2. Understand the query; A query begins with the system detecting the user’s preferred language. Phrasing is performed to assure that common phrases are detected and processed as a phrase rather than individual key words. The query is analyzed whether it is a general question, focuses on a problem, or is a highly specific question and narrow request.

3. Selective matching: Once the content and query have been thoroughly understood and refined, the system runs the query selectively against the appropriate index nodes. Content can be organized into topic-centric collections.

4. Present results in context: At a results page, the query is displayed as the user entered it. Query terms are highlighted in results, and results are clustered and categorized.

By grouping information in catalogue structures, it’s possible to navigate by searching. Automatic categorizing is placing the data in correct catalogues without the need for manual categorizing. At the scientific website scirus.com, FAST technology makes it possible to automatically categorize and place 100 mill scientific articles into the correct catalogues. I addition to this, the technology of FAST Data Search makes the eCommerce-experience easier, by enabling the customer to use the search engine as a tool for gathering product information, like customers can do at ibm.com and ebay.com using FAST technology.

Additional to the core modules, there are other add-on modules available, such as real-time filters. Real-time filter accepts high-rate data feeds, which match the data against stored queries, called “triggers”, and issues alerts that can be delivered through any messaging gateway such as fax, WAP/GPRS and e-mail.  Incoming content streams are matched to user profiles in real-time, and the same real-time content might be simultaneously indexed for search. The time between documents are added into a system and when they are searchable, is less than 10 minutes in the core FAST Data Search products. Users requiring faster indexing can add a Real-time search module, which reduces the time down to 90 seconds. For even more demanding customers there is the real-time instant access, which provides a sub-second indexing latency.  

3.0 What is different with FAST 

There are three unique parts of FAST’s data search technology that makes the technology competitive in the industry for search engines (exhibit 5)[8]:

1. Unique effective algorithms/compression
2. Linear scalability.

3. Integrated search and real time filter.  

1. Effective algorithms/compression

FAST’s effective algorithms/compression of data performs the search quickly and effectively. In addition to giving their customer’s added value in speed it also and decreases the need of data power, which again reduces costs. Increased data capacity by four times has been proven on existing hardware from traditional systems. In e-Commerce queries, the platforms have handled 100 percent increase without any additional hardware investments, still in sub-second performance.

2. Linear scalability

FAST’s architecture does not have an exponential increase of cost of increased system capacity. Traditional search engines technology will have an exponential increase of cost if your system increases from 10GB to 10TB. With FAST’s technology the increase of cost will be linear with the increase of capacity. This advantage is especially important for huge corporations with huge amounts of data.

3. Integrated real-time filter

All the big players in the industry have sub-second real-time solutions. FAST however is the only one that integrates both search and real-time filter into a single core technology platform. Unlike search, which compares incoming queries to historical data, real-time involves dynamic data, such as stock quotes, flowing into the real-time filter engine. This data is compared to an existing database of user triggers. If it matches the criteria of a particular query, the user is alerted and the data is served in real-time. The combination of search and real-time filter makes rapid customized content delivery possible. People can access pertinent information from huge amounts of historical and incoming data, which allows users to bridge the gap between information and action quicker. This offering gives FAST a competitive advantage.

3.1 FAST’s product line 

FAST’s product line consists of three different products developed through a similar core platform[9]. The unique products are adjusted and customized to the companies needs for handling huge amounts of information. When the market segments naturally, like it does in the enterprise search industry, it’s possible to design the search engine product line to match the needs. There are obviously different professionals with different needs in the IT-industry; some companies need information systems for single project use, while others are in need of wider deployments.

FAST Data Search Professional delivers capabilities with a low total cost of ownership, where the technology provides more queries per server and thereby reducing IT architecture, hardware and management costs. The product is bringing FAST’s best enterprise search capabilities to projects supposed to solve a specific project requirement. The product will search for information within a minimum amount of time, and at the same time maximizes the quality of the end user experience. The Professional software is aimed at departmental or single project use. (Price starts at $120 000)

FAST Data Search Enterprise is the product for data- and user-intensive environment, built on a highly scalable and flexible technology platform. The product is an integrated search platform developed for mission-critical information management solutions. The Enterprise search edition provides industry-wide and industrial capabilities and takes aim at larger, industry wide deployments.

FAST Data Search SST is the comprehensive real-time solution offered by FAST’s product line. Where information is mission-critical to the business and every minute counts in the decision process of the customers, the SST is the core of the real-time enterprise. The SST edition is geared towards the financial and media industries, which require access to real-time information. The Reuters Group uses the SST version and its contract stipulates that data has to be searchable within one second of being entered into the system.

4.0 The industry and its major competitors 

4.1 Industry 

FAST is competing in a fierce market where the search vendors chase the same sales opportunities, and some of the participants already have an installed base of customers (exhibit 3). FAST is recognizing Verity, Autonomy and Convera as their most important competitors. The participants of the search engine industry want to create a search platform that can support every type of user and the largest, most diverse content collections. Many of these competitors have already been in the market for a longer period of time, while FAST has had a later entry. While the industry for Web search has turned out to be quite homogenous, the industry for enterprise search is still very diverse, due to the differences in the present technology and customer base.

Gartner Group published, “Visionaries invade the 2003 Search Engine Magic Quadrant” the 24^th of April 2003. In competition with 21 other companies, FAST was ranked as the best company when evaluated in “competences of vision”, meaning they had the best core technology at present. FAST was ranked ahead of Verity, Autonomy and Convera, but was ranked behind Verity and Autonomy when it comes to access to customers. In 2002 FAST’s core technology was ranked as number 12. The improvement can be explained by FAST’s ability to sign important and difficult contracts with customers like IBM, Dell, FirstGov and Reed Elsevier. Gartner Group is known to be one of the best IT-analyst companies.

4.2 Market presence 

Verity and Autonomy are the large and established vendors with $100 and $50 mill in revenues in 2002, respectively. The two companies have the largest installed base of customers. Verity is the largest vendor when it comes to customers with above 1500 customers. Autonomy is the visionary participant, offering their products to 600 customers worldwide. According to their own statement, Autonomy only wants ambitious customers. FAST and Convera, the medium range companies when it comes to revenues, $46 mill (25% from Web-search) and $34 mill respectively, differs in customer base. Convera claims to have 750 installations, while FAST reports to have about 50-70 large customers. FAST is to a larger extent focusing on larger (customers with higher search-engine needs) customers than Convera.

4.3 Strategies 

The vendors see the same future of this market segment of enterprise search. Each intends to provide a platform that supports a wide array of users and content. Forrester Research believes this set up a fight to the death as each vendor attempts to capture customers and lock out competitors. Forrester Research believes the battle for search platform dominance depends on how the companies manage to match proposed features and their stated strategy and to find out which part of the market they are trying to cover. A crucial key factor is the competitors’ ability to make partnerships with other vendors in the software industry, like OEM-partners. Verity and Autonomy have developed extensive relationships. Another key factor is the breadth of users and content that each can participant can cover. Google can find the most relevant sites on the Web, but on the other hand, their engine will have a hard time satisfying intelligent agents critical to their searches.

4.4 The major competitors 

4.4.1 Autonomy 

Autonomy is a spin off of Neurodynamics ltd. and was established as an independent group in 1996 to commercialize its advanced concept of searching technology beyond the security and defense industries[11]. Autonomy started on with recognizing patterns for fingerprints. This technology of pattern matching techniques has been brought along into the search engine industry. Although Autonomy has faced a tremendous growth in their industry, the company is still small when it comes to organizational matters. Today, about 200 employees are working for Autonomy worldwide.

Autonomy’s strengths lies in a unique combination of technologies that employs advanced pattern-matching techniques, utilizing Bayesian inference (the rarest words in a document are the most important) that together with Claude Shannon’s principles of information theory[12]. Traditional search engines use tags like meta-tags to improve the results when searching for keywords. Autonomy, on the other hand, claim they provide full automation for operating unstructured information by using a software independent of linguistics and language. Autonomy’s technology forms an understanding of the actual content of any type of information, thereby enabling Integration Through Understanding (ITU), which allows applications to communicate with each other without any manual effort involved in setting up complicated connectors or the use of Meta data. The advantage of this software is the fact that the software does not have to be updated for new words and phrases like traditional search engines have to do. Autonomy’s technology is modular and can accommodate 200 different data types, processing text, voice and video sources. Autonomy's infrastructure technology enables computers to form an understanding of a piece of text, Web pages, e-mails, voice, documents and people automatically. According to a report by Giga Information Group in January 2003, pattern matching and recognizing technologies will replace rule-based tools. This is because the technologies allow applications to create more complex associations between content, tools and people. The technologies can enable information to be delivered in context to what employees are working on, creating an adaptive environment.

Often the user doesn’t know the extent of searchable information. Autonomy’s real-time agents are supposed to help finding relevant information by establishing customer profiles. The registered profile will automatically receive available and relevant information, found by a set of agents searching the databases for information.

Autonomy counts 600 customers worldwide, ranging from telecom industry to media companies. With the focus on the automated agents working in real-time, Autonomy is probably the most developed participant in this diverse market. The company is strongly focusing on the licensed products; most of Autonomy’s revenues originate from re-sales through licensed products. These re-sales are primarily done through partners, like BEA Systems, CGEY and Novell, which can implement search technology into their customer’s solutions.

4.4.2 Verity 

Verity is recognized as the market leader in information retrieval software with over a decade of experience in information retrieval and portal technology.

Founded in 1988, the company recognized early on that knowledge is power and developed one of the first full-text search solutions[13]. Verity counts more than 1500 businesses worldwide in virtually all industries among its customers. They include nine of the Fortune 10, 80% of the Fortune 50, and 66% of the Fortune 500[14]. Verity’s technology also powers some of the most extensively implemented business applications in the world, including Adobe, Documentum, PeopleSoft, and TIBCO. In November 2002, Verity acquired Inktomi’s enterprise part of business. Through this acquisition Verity got 2500 smaller departmental customers from Inktomi. The acquisition brought Verity a stronger alignment with the Enterprise end of the information management spectrum, but significantly, it provided Verity with access to the tantalizing market segment known as "the low-end”[15]. Verity is hoping they can upgrade these companies systems as their demands are growing. The different segments can be highlighted by average deal-size, which were $150 000 for Verity and $18 000 for Inktomi’s enterprise part.

Verity’s core technology is known as a strong Boolean base with keywords and thesauruses to yield precise, controlled searching. This is a traditional keyword technology built on advanced algorithms and linguistic analysis technology. This technology is dependent on a good “keyword” description to perform a good search. In comparison to Autonomy who has a search system that can learn what users consider relevant, simplifying and improving search results over time. This disadvantage though is that it requires well-selected samples to learn from.

To summarize one can say that Verity uses a well-proved and well-developed technology, which they have developed to contain several advanced Knowledge Management solutions for enterprises. Furthermore, Verity has a big installed base of customers and can take an advantage of their lock-ins.

4.4.3 Convera 

Convera was established in 2000 through the merger of the former Excalibur Technologies Corporation and Intel Corporation’s Interactive Media Services division. Convera designs, develops, markets, implements and supports search, retrieval and categorization software including text, audio, images and video irrespective of language[16]. Its core technologies include: advanced computational linguistics and semantic networking that leverage lexical knowledge using built-in knowledge bases to search not only for specific word meanings, but also for related terms and concepts[17]; Adaptive Pattern Recognition Processing (“APRP”) that identifies patterns in digital data, providing the capability to build content-based analysis and retrieval applications for any type of digital information, not too different from Autonomy’s pattern matching technology. In combination, these core technologies form the foundation Convera builds its products on.

Convera has about 260 employees and a base of 750 customers ranging from large to smaller companies. Convera has as Autonomy their greatest strength in content management as accessing, organizing and utilizing enterprise data. This is enabling customers to optimize their data. When it comes to normal search it doesn’t seem as strong as a keyword-based engine.

5.0 Who are the customers and why do they choose FAST? 

During the last years enterprise search has gone from a “nice to have” to a “must have” for many companies. According to a report made by the Yankee group (The Benefits of Effective Search Technology, 2002), 

 Search has become the second most important source of information after communication in major companies. 

FAST’s customers are at the moment mostly big companies with incredible amounts of data stored and where the need for an effective search engine is crucial for their business. Examples of FAST customers are IBM, Dell, eBay, CareerBuilder, Reed Elsevier, Reuters, Lycos, Spray Networks, Telenor Mobile, Banca IMI, FirstGov and Ericsson.

5.1 Why companies choose FAST 

We have looked at some companies to see why they chose FAST over their competitors.

5.1.1 IBM 

IBM has one of the largest websites on Internet, which serve customers in many different ways. Customers can check for information, they can shop new products, download drivers and updates etc. A good way to seek information on a website is to search for it, especially at big websites like IBM’s. For this reason, IBM is looking at their enterprise search as one of the critical factors for user- friendliness on their website. At the moment the search engine at IBM.com handles an average of 135 000 queries per day[18]. In a situation like this FAST claim that they have reduced their cost, increased capacity and reduced serving time for the customer at IBM.com. 

5.1.2 REUTERS[19]  

Reuters is the world’s largest international news and television agency, publishing over 30 000 headlines daily in more than 26 languages. The company serves more than 560 000 users in 52 000 locations worldwide.

Reuters desired to provide their customers with an enhanced capacity to deliver them crucial business information over the Internet.

They required their new search engine to be able to:

1. Handle Reuters content, brokered content or any news feed.

2. Provide for both personalized push (alerts) and pull (full text search)

3. Ability to scale real time alerting and filtering to hundreds of thousands of subscribers

4. Ensure sub second indexing latency

After thoroughly research they found the preferable solution from FAST to fit their requirements best. They chose FAST because the company was able to deliver both push (real-time subscriber information) and pull (historical, static data) information to their customers in an economic and effective way. FAST with its linear scalability, and its combined core technology with both real time and static search simultaneously, was cost effective and efficient.

5.1.3 FirstGov 

FirstGov is the U.S. Government’s official portal to all government information services. The web site was launched in September 2000, and redesigned in 2002. During the redesign FAST was contacted. The first solution mostly returned web pages (Html.files etc) on their web site, now they also wanted the search engine to find Word files, PDF-files, database records and several other file-formats (Excel, PowerPoint, XML, plain text etc). In addition to versatility the engine also needed to be able to index up to 200 million pages and handle queries from approximately 20 mill visitors a year.

The competition of the contract was fierce and FAST with its partner AT&T presented the only solution that either met or exceeded the requirements from FirstGov. Again FAST showed their good capabilities for large websites.[20]

5.2 Why companies do not choose FAST 

The companies in the enterprise search and information sharing industry seem very similar at first sight. However, all of them deliver specific solutions the other ones don’t have. This makes the companies to some extent niche players within different industries and different companies with various needs. Therefore, there are many reasons why companies chose different enterprise solutions.

5.2.1 Norsk Hydro 

Hydro chose Autonomy’s solution when they looked for a worldwide Knowledge management system (enterprise-search and information-sharing software) for their employees. Autonomy claims they have a unique fingerprint agent making it possible to automate operations in the unstructured information base. This technology works in a way that the employees register in the system with their interests, job-position etc, making their own fingerprint. The agent from Autonomy then recalls this fingerprint and automatically sends the relevant information to the right users when there is a match between the documents’ fingerprint and the users’ fingerprint. This means that the users don’t need to search for the information, instead it is automatically sent to them. The agent is at the same time showing the users which colleague in the organization which is interested in the same information as themselves, making it possible to contact relevant people across departments, divisions and countries. This increases the information sharing within the organization making it more efficient.[21]

Norsk Hydro had a vision of improved information sharing and felt Autonomy had the best solution with their automatic agent, which automatically gives the right employees the right information from the right relevant sources.

5.3 What customers want 

The information search industry is very complex and the solutions are often tailor made and customized for different customers. For this reason, a company is often chosen on the level of individualization the different companies can give. Customers have different needs. Some companies like Opplysningen 1881, a Norwegian directory enquiry service, do not need a smart agent, only an effective search engine, while others needs an advanced software-package reaching far beyond a single search engine. These different needs make it important with individualization for the companies. You can say that customers who are willing to pay a lot for these services are the companies who can gain a competitive advantage to their competitors by installing a unique knowledge management system[22].

Customers are often forced into long-term relationships because of the extensive process of implementing the system, resulting in lock-in effects. Customers are therefore interested in solutions that work across different platforms making it possible with further improvements and expansions in the future. Customers know it’s hard to avoid these lock-in effects, and they are at least looking for some benefits or sweeteners up front as a compensation for lock-ins. As a supplier in the information industry, the vendors have to think ahead of the entire lock-in cycle as they negotiate. The vendors have to be creative in what they seek; service and support guarantees, free upgrades for some period of time, or most favoured customer treatment[23]. It will be crucial to cultivate the influential and most profitable customers, and focus on building switching cost among these customers. When evaluating new customers, the vendors have to estimate how much each customer’s switching costs are, and determine potential revenues to earn from the customer and thus the potential investment of acquiring the customer from the competitors. Switching costs are the norm in information industries. It’s impossible to compete effectively in the information economy unless you know how to identify, measure, and understand switching costs and map strategy accordingly

When we had a closer look at different customers in the different competing companies we found out that some of the companies actually are customers of more than one of the enterprise search companies. Reuters and Ericsson have a relationship with both FAST and Autonomy. This underlines the specific needs some companies have. Ericsson has chosen a solution from Autonomy that FAST could not deliver, but at the same time chosen a solution from FAST that they felt was better than the solution from Autonomy. This shows that this is not a very mature market. None of the companies are really a one stop shopping if you want the best worldwide solution for information search and sharing within an organization.

Another issue is the customer-focus among the different players. As Dick Senmark, CIO at Volvo, puts it well in an article in the December 2002 issue of CIO Magazine[24], "The search engine industry and the research community alike often fail to acknowledge that intranets are not just downscaled versions of the Internet, but are instead a whole different environment in terms of both content and culture. Intranet owners must start to ask for search products specifically designed for intranet use, and more important, that fit the work environment at their company." This can also be a key element for the future. Managers will in the future probably concentrate less on that the best search engine is the one who comes up with most hits, the so-called “Google-syndrome”. In the future maybe managers will go away from “more is better” and start focusing on the benefits of being able to sort by date, author, subject relevance, and many other parameters. Such a search gives fewer hits but higher relevancy for the users.  

The companies that most likely will win the battle of the customers in the future are the ones that can differentiate their products and easily individualize their products, making them unique and give their clients an advantage over their competitors. At the same time the companies need to develop their technology and expand their solution into new segments as the future technology develops and thus moves more towards one stop shopping.

6.0 Organizational Issues 

FAST is relatively small company with an organization of approximately 160 employees. One of their main objectives is to keep their cost down. In addition, they have little interest of organizational growth. FAST as an entrepreneur has traditionally focused on technology development and less on sales and marketing. FAST is competing in an environment where their main competitors have been in business for a longer time than FAST. FAST’s major problem in the fight for market share has been, to a certain extent, a poor effort in marketing their products, as a consequence of facing the problems and challenges of running a technology-driven company. Still, development and R&D is the main focus in the organization, where most of the employees are developers.

However, in a situation with growth in markets and revenues, the organization needs to use increased resources on sales and marketing. By increasing collaboration with distributors like CGEY and IBM, FAST has been able to keep on performing better without organizational growth. At the same time they have managed to keep up with their technology focus within the company. This strategy fits FAST and a technology-focused company will probably be more attractive for competent developers than a large organization with a large sales-implementation staff.  

Companies in the CRM and ERP-business are potential partners for indirect sales[25]. SAP and PeopleSoft are potential partners in the ERP-business, possible to implement search technology into their platforms. Throughout the coming year FAST has a defined goal of 40% of sales through distributors and partnerships, compared to 20% in indirect sales in 2002. These relationships will hopefully increase FAST’s margins and profits, when partners and distributors are doing the implementation. During 2002 FAST managed to sign contracts with new distributors, new customers for direct sales, in addition to renewing existing contracts with profitable customers like Reuters Group and Reed Elsevier. Through their strategy, FAST is able to continue with their core activity, which is development, at the same time as reducing their fixed costs and thereby making the company more flexible for future changes in demand or technology.

7.0 Market opportunities and obstacles 

The search and information sharing industry is an industry with potential to increase the efficiency of information technology in general. The total search market is increasing year by year; according to FAST the total search market is increasing by approximately 20% every year. The growth is a consequence of continuously increases in the amount of data in intranets as well as on the Internet. In addition, big companies are starting to focus more on information sharing/Knowledge management to give increased access to information for employees as well as their customers. In a more globalized world where companies have activities all over the world, the challenge of knowledge management within the organization is increasing. It is important to have systems making it possible to utilize synergies locally as well as globally. Utilizing these potential synergies has created an industry of search engine vendors. Information is costly to produce, but it’s cheap to reproduce. Large up-front sunk cost, minimal capacity constraints, and low incremental cost has allowed certain participants to be profitable in this industry.

As described above, the market for enterprise-search is currently large corporations, multinationals, governments etc with large amount of data and a huge organization where information sharing is crucial for their business. In the business environment of tomorrow it is expected that also medium and after a while small sized firms, which currently use less advanced search engines attached to and developed from companies like Oracle and Microsoft, will need more advanced enterprise search solutions. The entrance of medium and small size companies will open a huge attractive market. The question is how and if FAST has the abilities to exploit these opportunities. Should FAST continue to focus on cultivating huge companies, while at the same time trying to steal customers from their competitors in the existing market, or try to be an innovator in these new segments?

Among most companies it is considered difficult and expensive to implement a search solution. This is one of the reasons why many medium-sized companies still use simple search solutions from Oracle and Microsoft; this is convenient for companies with fewer resources. If FAST enters this market they have a challenge to make their systems more compatible and thus easier and cheaper to implement for smaller companies. Collaboration with ERP and CRM companies like SAP and PeopleSoft, or middleware vendors like BEA Systems can be a key element to meet this challenge. Users are generally not interested in where search products come from, as long as it works well. If FAST’s solution is implemented in for example SAP and thus fully integrated and installed together with SAP, it would make the product more easily implemented. At the same time they would get access to a huge base of companies currently using SAP systems. An exclusive agreement with SAP would for this reason boosted sales in all segments. At the same time it would have increased their opportunities to fight lower segment participants like Microsoft. However, the search industry is still developing and the ERP/CRM -companies are afraid to make an exclusive agreement with a failing technology. Consequently, such exclusive contracts are of crucial impact in this competitive and developing industry, and the fight for these future contracts will be fierce among participants of the enterprise search industry.  

Sales are currently concentrated around the “high-end” market, and growth in this segment would most likely lead to a decline in the coming years as it becomes saturated. Such a situation will increase the competition in the segment and thus the probability for a price war. To keep focus only on the “high-end” segment would therefore involve a high innovation/differentiation strategy making it possible to sell at higher prices. Otherwise, the company might be pushed in the same way as in the web search market. In this market FAST experienced a stop in market-growth, leading to increased competition on price, which eventually led to the mentioned consolidations in the web search industry.

Increased possibilities of more advanced and unique customized solutions will probably be more important in the “high-end” market. Customers will probably demand more from the solutions when it comes to issues like security and confidentiality for internal documents as well as advanced agent solutions for knowledge management. Simultaneously, more standardized and thus less expensive solutions will probably be a success factor for the medium and small sized companies. Such a solution can easier be implemented and will be important to increase the total market for advanced enterprise searches and information sharing market. Thus differentiation and standardization are aspects for the future.

8.0 FAST’s future? 

FAST has experienced a huge growth and development in the industry since they started up. FAST has during the last year done a lot of successful decisions and emerged into this industry as one of the major players. However, the industry is still developing at a rapid growth with fierce competition, consequently no one knows how the industry will be in the coming years.

Where should FAST go from where they are today? How can FAST further strengthen their brand in the market and improve their market position? How can they exploit their advantages and further capitalize on its enterprise-search capabilities? How can they best face the challenge of growth within their organization, by organizational growth or through partners?  

9.0 Exhibits 

Exhibit. 8

FAST’s Management Team[30] 

John M. Lervik, Ph.D., serves as the Chief Executive Officer and is a co-founder of FAST. Dr. Lervik served as the Company's Chief Technology Officer from 1997 to September 2001 overlooking all of the company research and product development activities. Dr. Lervik holds a Ph.D. from the Norwegian University of Science and Technology, and was awarded the best overall PhD at NTNU in 1996/97.

Ali I. Riaz serves as the Chief Operating Officer and Chief Financial Officer with the dual role of managing the financial and corporate function strategies. Mr. Riaz previously served as the Vice-President, Strategy and Corporate Development since October 2000. Prior to joining FAST, Mr. Riaz held executive positions with Computer Sciences Corporation (CSC), including Vice President for Finance and Administration for Europe, Middle East and Africa. Mr. Riaz has also held several management positions at Novartis Pharmaceuticals, including Head of Planning, Information and Control and Head of Global Marketing Controlling. Mr. Riaz holds an MBA degree from Loyola Marymount University in Los Angeles, California.

Bjørn Olstad, Ph.D., serves as the Chief Technology Officer. Before joining the Company, Dr. Olstad held key positions within General Electric Medical Systems, including Director of Research and Development for Cardiac Ultrasound. He has served as a professor in computer science at the Norwegian University of Science and Technology (NTNU), where he was awarded the youngest professorship ever.

David M. Burns serves as the Executive Vice President for Business Development, and is responsible for major alliances and channel development. He has been with FAST since 1998 when he launched the company’s North American division and served as President of our U.S. operating subsidiary Fast Search & Transfer, Inc. Prior to joining FAST, Mr. Burns served as Director of Product Marketing for Lycos, Inc., and Vice President of Marketing for NobleNet, Inc. He holds a dual B.Sc. in computer science and electrical engineering from the University of Connecticut and an MBA from Babson College.

FAST Board of Directors

Thomas J. Fussell has served as Chairman of our board of directors since July 1997. He also serves as executive chairman of the board of directors of Opticom ASA and as a director of Gaelic Resources Plc, an oil investment company listed on the London and Dublin Stock Exchanges, and Philippine Hoteliers, Inc. in the Philippines. Since 1992, Mr. Fussell has served as managing director of Charles Street Securities Limited, an investment and consultancy firm that he co-founded with Mr. Robert Keith specializing in corporate advice to energy and high technology firms. Before 1992, Mr. Fussell worked as a research analyst in natural resources and corporate finance in London, specializing in emerging markets.

Robert Keith has served as a Director since July 1997. He also serves as managing director of Opticom ASA. From 1995 to 1998, he served as a director of Eidos Plc, a computer game developer and high technology research and development company listed on the London Stock Exchange and the Nasdaq National Market. In 1992, he co-founded Charles Street Securities Limited with Mr. Fussell. Before 1992, Mr. Keith worked as a research analyst, salesman and corporate financier at Carnegie International Ltd., London.

Hans Gude Gudesen has served as a Director since July 1997. He also serves as a director of Opticom ASA, a company he founded, and as a director of Thin Film Electronics ASA, a subsidiary of Opticom. Mr. Gudesen has coordinated a broad range of research and development projects in the fields of retrieval and data compression, which have resulted in the formation of a number of research and development companies, such as Opticom ASA in Norway and Smart Materials S.A., a Belgium corporation.

Øystein Tvenge has served as a Director since June 2001. He was born in Oslo, graduated from Norges Markedhøyskole and is now based in Oslo, with his main business focus being on real estate development and the software industry. Mr Tvenge has also been involved with media enterprises and various start ups and innovative technology companies. In particular, Mr. Tvenge was the founder of Norwegian Yellow Pages (since sold to Telenor ASA), co-founder of Agresso Group ASA (now Unit4- Agresso N.V. ) , a leading European business software provider, and he is a member of that company's supervisory board. He is also co- founder and board member of Investra ASA, an OSE listed investment company, which is controlled by Mr. Tvenge and his brother. Mr. Tvenge also serves as chairman of the board of Vanadis ASA (ex. Fastweb ASA), an unlisted company specializing in web hosting services and internet based media enterprises and as executive chairman of Concept eSolutions ASA, a Norwegian unlisted software company, with its principal operations in France, Spain and Italy.

Bryan A. Burdick has served as a Director since February 2002. He is the Vice President of Portal Services, for TerraLycos U.S., the leading global Internet portal, and manages six core business units including Search, Access & Communications, Multimedia, Personalization, Communities and E-Commerce Services. Prior to joining TerraLycos, Mr. Burdick worked at InfoSpace, Inc. (formerly Go2Net, Inc.) first as Vice President, Managing Director of Go2Net's Silicon Investor business unit and then as SVP, Broadband Products where he launched several new businesses in the broadband consumer, SME and interactive television markets. A graduate of the Wharton School of Business, Mr. Burdick has more than twelve years experience

• [1] Information found from: Press releases, newspapers, and interviews with employees in FAST Search and Transfer ASA (Contact person: Camilla Myhre).
• [2] iMarkedet.no, 2000