The Business Web in Russia: Usability for the Western User

An investigation of the part of the World Wide Web (Web) devoted to business purposes in Russia was conducted. The goal was to estimate usability of this Web for the Western user by focusing on the following dimensions: type of businesses for which Web sites are used, language used on sites, communication capability of sites, ordering capability of sites, and response time of site servers. The findings suggest that the Russian Web is more usable in terms of language used and communication capability than ordering capability and response time. Neverthless, the emergence and the state of the Russian Web signal significant telecommunication and market developments is Russia.

1. Introduction

The purpose of this study is to explore business-related sites on the Russian Web. The importance of the Russian business-related Web stems in part from the importance of Russia itself, being the major country to emerge form the Soviet Union in 1991. In spite of its current economic and social disarray, Russia will likely continue to play an important international role. With this in mind, there is value in studying the Russian business Web, in that it can be seen as a mirror for both telecommunications and broader economic developments in the country. The implosion of the Soviet Union resulted in the collapse of the voluntaristic, party-government controlled economy. As a consequence, Russia began the transition to a market economy. The telecommunications sector, more specifically computer networks, has been one of the pioneering areas for market reforms and is characterized by significant foreign investment, internal competition and elements of deregulation. Computer networking, in turn, has supported the development of other businesses by providing new channels of domestic communication (Travica & Hogan, 1992).

The genesis of the business Web may signal a new evolutionary phase in both Russian computer networks and economic development. Networks are entering a more mature phase with the spread of the TCP/IP (Internet) standard, which appears to be a de facto international standard, while the economy is opening up to the outside world. Both developments are mirrored in the growth of the business Web (see Cronin and McKim, 1996).

Our study has number of limitations. One has to do with the novelty of the phenomenon studied. To illustrate, the Internet was introduced in Russia in 1991, although OSI-ISO computer networks do have a longer history in the country (see Travica & Hogan, 1992). As a result, there is almost no previous research on the business Web in Russia. Another limitation is the dynamic, "real-time" character of the study. Consequently, we have defined our goals modestly: (1) describe the Russian business Web; and (2) test several aspects of the Web from a Western user's perspective.

2. Methodology

2.1 Dimensions Studied

2.2 The Sample

Once the list of Web sites was identified, an effort was made to select those related to business. Business sites were considered those whose owners were non-governmental firms (e.g., telecommunications providers, software and hardware vendors, banks, and online magazines) and/or which provided information related to business activities (business law, news, bids and offers, banking, firm addresses, price lists, advertisements). Negative conceptualization also helped in the classification: a business-related site was considered to be one that was not related to the academy, research or government. After several iterations, we identified 54 business-related Web sites in Russia, and estimated that this number represented at least 80% of the population. (For example, we ran a check against hot-lists stored on Russian Web sites only to identify some additional telecommunications sites.) Determining an exact number of sites was impossible for the reasons mentioned above. The inaccessibility of certain sites was also a problem. Since the telecommunications providers were dominant in the sampling frame, weights were assigned to less frequent types of organizations to increase their chance of being selected; a random sample of 35 sites (65% of the sampling frame) was subsequently selected. The sample drawn appears to be representative in terms of kinds of organizations and their relative proportions.

2.3 Data Collection and Analysis

All the dimensions studied are categorical, except for response time. Since most of the categories are straightforward (e.g, the language used), no particular data analysis technique was necessary for determining their values. Type of business, however, required a more systematic approach which employed successive iterations and revisions of coding similar to the Glasser and Strauss (1976) grounded theory approach. Although ordinary declarations of a site's purpose, which were found on almost all sites, should have made it easy to determine this category, it was only useful when the organization's business was unambiguous and unidimensional (e.g., banks, brokerages, online magazines). In other cases this was not the case; for example, although each telecommunications provider would have the words "telecommunication" or "communications" in the title of its home page, a closer look revealed that some also provided business-related information. Another problem was how to determine the primary business according to the categorization which was used in the study; for example, we classified as business-information providers those sites whose primary offerings could be understood as business-related data. Yet another problem arose with sites that provided mixed clues about their business. For example, the Angar site, which describes itself as "experimental," contains technical data about foreign and domestic computer equipment, which might have qualified it as an advertiser; however, there is no price data provided (but a hypertext link to "Price List" which starts email for inquiring about prices). Also missing were data on firms and other content that constitutes an ad. We classified Angar as a hardware vendor, because some textual clues suggested that the company was indeed selling the hardware described in its pages.

3. Findings

• a geographical skew: the highest density is in the Moscow and St. Petersburg regions;
  
• a state of heavy construction: around 30% of the sites in our sample were underdeveloped (indications are either developers' polite excuses or error messages like "The page not found"); and
  
• the dominance of telecommunications companies.
  

3.1 The Dominance of Telecommunications Providers and the "Russian Internet"

Table 1 - Businesses Profiles of Web Sites

Type of Business	Frequency
Telecommunications Provider	14
Business Information Provider	12
Computer Software/Hardware Vendor	6
Advertiser	4
Online Magazine	2
Systems Services Provider	2
Business Services Provider	1
Bank	1
Blood Bank	1
Brokerage	1
Finance	1
Matchmaking	1
Total	46

Note: The number of "Types of Business" is greater than the number of sites visited, because some of the sites were coded with more than one kind of business

In Russia, the Relcom (the name comes from "Reliable Communications") is commonly called the "Russian Internet," but officially it goes by "Demos/Relcom network." The popular name reflects the fact that the Relcom was the forerunner of Internet connectivity, when it linked the former Soviet Union with the global Internet in 1991. Since then, the Relcom has developed into the main TCP/IP network in Russia. The official name of the network denotes that Demos Co. was one of the founders of the Relcom. Demos was established in 1989 as a software and network company with the primary purpose of creating an operating system for mainframes and workstations. ("Demos" is an abbreviation of operating system name: Dialogue Unified Mobile Operating System.) In 1990, Demos was given the task of developing a small internal computer network for the elite Kurchatov Institute of Atomic Energy (KIAE; now "Kurchatovsky Institute") and other scientific organizations in Moscow. KIAE thus became a co-founder of the Relcom network. (Relcom, Demos, 1995)

The software that defined the Relcom in the beginning was the UUCP suite of protocols (Unix to Unix CoPy) which provided slower data transfer than the Internet's TCP/IP suite. The Relcom also possessed an international link--a dial-up UUCP link to an EUnet host in Finland was established in August 1990. This link turned out to be an asset that helped the Relcom overcome its limited role of serving a handful of Moscow-based scientific users. During the attempted coup d'etat in August of 1991, the Relcom operators were very active in informing the world about the events in Moscow, and Russia about the world's reaction to the coup (see Travica & Hogan, 1992). The failure of the coup created an opportunity to expand. In September 1991, the Relcom obtained the Internet country domain .su. Since then it has been evolving along with its foreign counterpart EUnet, which has developed into the largest provider of Internet services in Europe (Internationally, Relcom is recognized as the EUnet/Relcom Corporation).

The current offerings of the Relcom include a set of standard Internet services (telnet, ftp, network news), and the high bandwidth (42 Mbit) services like video-teleconferencing and Web (these services belong to a higher price category). The old UUCP services (email, file transfer, telefax, teletype and telex) are preserved, and comprise a lower price category. In addition, gateways to networks are based on the OSI architecture (e.g., the Infotel and the MMTel networks based on the X.25 protocol) (Relcom, 1995). Organizational evolution accompanied the technological development. The network is currently operated by the Relcom Joint Stock Co. Growth indicators reveal a doubling of the user population on an annual basis -- in Fall of 1995, Relcom served 200,000 users (ibid.), in comparison to 30,000 at the start of 1992 (Travica & Hogan, 1992). The new users are individuals as well as organizations, such as banks, exchanges, industrial enterprises, information agencies and government bodies. The Relcom has thus become not only the largest computer network in Russia, but also in the entire territory of the former USSR.

MARK-ITT Communications Company is an example of a successful company which has close business relations with Relcom Joint Stock Co. It was established in April 1992, with the mission of developing the EUnet/Relcom network services and creating its own information and communication services in the Udmurt Republic in the Russian federation. In 1993, MARK-ITT's clientele was growing at a rate of 5-10% monthly. Growth slowed down after that, while the company continued to increase connectivity and the range of services. For example, in November 1995 it offered the following services: the transnational EUnet/Relcom node in Budapest, links to several cities/regions that resulted from the cooperation with providers of local communications, packet-switched services resulting from the cooperation with the Rospak Joint Stock Venture, and fax services.

It is interesting to note that the MARK-ITT business has been growing in the same way the Relcom's has: connectivity is being extended over a new territory by establishing business cooperation with providers of local communication services. This indeed represents the model according to which the deregulated development of computer networks is unfolding in Russia.

3.2 Information Providers and Other Businesses

The advertiser category refers to the sites that are used for advertising a second party's offerings. While the dominance of the telecommunications and information providers is hardly surprising, the relative absence of advertisers is (see Table 1). The number of ads per advertiser site is also small -- between two and five. The ads usually consist of a firm name and address; a firm logo accompanies the text in some. In a few of the more elaborate ads, the hyperlinked pages were comprised of price lists. The use of color and attractive backgrounds is very limited, while menus in the form of image maps were found at only one site.

It is interesting to note that advertising in the Russian Web can sometimes be whimsical. For example, there was a link at several sites pointing to the home page of a brewery in Germany. In the brewery's Web page a user would be offered a deal: include the brewery's URL in your Web page and receive a six-pack of beer. During our investigation, the brewery introduced a new stipulation: a six-pack could be won only by a person who would be the first in his city to include the brewery's URL in his page. This kind of advertising may reflect the country's relative lack of marketing savoir faire (see Discussion section).

The advertiser category excluded situations in which a company advertised its own services or products. These differ considerably from their Western counterparts. For example, the user can learn much about a company's history and projects, but a price list may be difficult to obtain. It is perhaps assumed that an interested user would send electronic mail to the site maintainer, or find some other way of obtaining price information; this apparently differs from the Western advertising philosophy that aims to make it as simple as possible for the customer.

Other kinds of businesses in our sample include some of the largest enterprises of their kind (e.g., Roscredit is the largest private bank in Russia), or industry pioneers (e.g., the blood bank Sanguis). Moreover, some sites demonstrate a mix of business types. For example, the Dux company not only provides access to the Internet, system design services, and software products, but also offers the prospective Web page owners the server space for rent. It is possible that, in the chaotic processes of market creation in Russia, this multiple orientation is rather the matter of opportunities and spontaneity than of clearly defined missions and strategies.

3.3 Language

There is also a Mix category in Table 2 that indicates a failure to provide consistently mono- or bi-lingual sites. For example, a user who pays a visit to the site of a major telecommunications provider must often navigate through Russian and English pages that switch without warning.

For the user that chooses Russian, a convenient function of fetching decoding software is almost universally available. By clicking on appropriate labels, the user can start ftp and download the software which displays Russian Cyrillic in the MS-DOS, MS Windows or Unix environments. The user can install the Russian fonts, say, in a Windows environment, and then start them at any time before or during the browsing of the Russian Web. The decoding software automatically converts strings of International Latin characters which are created by encoding the Russian Cyrillic at the source point by appropriate encoding software, and are used for transfer purposes. Since these characters use the extended ASCII code, the use of decoding software at the receiving end does not affect the English set of Latin characters. Consequently, the user can switch from Russian to English pages without leaving the Russian font mode. The most popular encoding-decoding system for MS-Windows is the so-called KOI-8, which works satisfactorily under Netscape in any MS-Windows environment.

It is interesting that at the time of data collection English was the only foreign language found on the Russian Web, especially since both German and English are officially accepted as business languages in Russia. Also, some price lists are expressed in Deutschmarks (the rest are in Russian rubles and U.S. dollars).

Table 2 : Language at Web Sites

Language	Sites
Russian	4
English	6
Bilingual	23
Mix	2
Total	35

3.4 Communication Functions

Table 3 - Communication Functions

Communication with Site Maintainer

Yes	No	Uncertain	Total
26	4	5	35

Ordering Capability

Yes	No
2	33

The data in Table 3 suggest that most of the servers investigated do provide communication channels with the server maintainers (usually the persons in charge of the technical aspects of the sites). This communication function is typically accessible via a clickable' email address which starts email function. (We did not consider mere provision of e-mail or postal address a communication channel.) Furthermore, about 10% of the servers did not provide the communication capability, while for 13% of the servers it was not possible to determine if the capability existed.

The other communication function, automatic ordering, was offered on only two sites (less than 1%). This may have to do with the state of marketing in a country that quite recently began moving toward a market economy.

3.5 Server Responsiveness

Table 4 - Server Responsiveness

Measure	Finding	Site	%
Min. individual Response Time (tr)	2sec	MNTS	-
Max. individual tr	105sec	Angar	-
Min. average tr	5sec	Roscredit	-
Max. average tr	50sec	Telsib	-
A Speed (tr 0-15 sec)	18 servers	-	56%
B Speed (tr 16-30 sec)	8 servers	-	25%
C Speed (tr 31-60 sec)	6 servers	-	19%
Consistent tr (Standard deviation of individual tr <= 4)	8 servers	Cronyx, AMT, Surnet, Roscredit, Relcom, Demos, ICE, Garant-Siberia	25%
Response Reliability (Servers responding each time) Each Time	22 servers	-	69%

Is there a relationship between the length of response time and consistency? Correlational analysis suggests that such a relationship exists. Specifically, correlating server average response times with standard deviations of the server average response time gave a Pearson r=0.87 (p<0.0001). This suggests that servers which respond more slowly are also likely to exhibit less consistency in response times in successive access trials. The reverse is also true: the more quickly responding servers are more likely to respond within ranges that do not vary significantly from one access trial to another.

3.6 Other Observations