![[© 2010 • ISSN 1420-3693 • www.localization.org]](/globalizationinsider/images/masthead2.gif)
|
Market Survey of Multilingual Websites
Steve McClure scans the statistics to reveal the patterns and lessons of website globalization in this preview from the forthcoming issue of the LISA Newsletter. The Devil is in the Details
The software markets for ebusiness globalization software, multilingual content management, and cross language solutions are all emerging markets. Although some larger vendors like IBM are pursuing these markets, they tend to be characterized by smaller start-ups. Fledgling companies frequently spend a disproportionate amount of time evangelizing as they attempt to find the most receptive audience for their hot new technology. Any information that directs start-ups to the better prospects in the market (e.g., focusing on specific niches) increases their chances of success and maximizes the utilization of scarce marketing and sales resources. In this article, we analyze the characteristics of multilingual websites as potential targets, using data from a recent worldwide survey, IDC's eWorld 2001 Survey. eWorld 2001 Survey
IDC's eWorld 2001 Survey focuses on ebusiness adoption trends across 27 countries around the globe. The survey effort took place from January through March 2001, although the Eastern European country surveys were conducted several months later. The sampling unit of this survey is an establishment, or site, which is defined as one physical location of an organization. The sample of business sites consisted of 12,698 respondents across 10 vertical markets and five size classes. Eligible respondents were the individuals at each site best qualified to speak about the site's overall computing activities. The results may be generalized only to sites with Web access, not all sites. Interviews were conducted via telephone. This bulletin uses survey results weighted by employment and IT spending. The employment-weighting scheme reflects the distribution of employees at sites across size and industry, and the IT spending weight reflects patterns of IT expenditure at sites across size and industry. Interpreting the Survey ResultsThe results may be generalized only to sites with Web access, not all sites, because if a location did not have Web access the interview was terminated. A second proviso is that many of the results discussed in this bulletin relate to the subset of respondents whose company had a website accessible to the public. It is possible for a location to have Web access but not have a website accessible to the public. In fact, in the eWorld 2001 Survey, three out of every eight respondents did not have a website accessible to the public. Not all of the questions were asked in every regional survey. So in our discussion of Table 6, we limit our remarks to just the United States and Canada. Multilingual websitesWhen prospecting for buyers of software, where better to look than in categories of companies that already support multiple languages on their websites. The survey collected data on support for 16 languages plus "other" (see Table 1). From this data, we can also draw conclusions about whether sites are monolingual or multilingual. By doing so, we are able to analyze the characteristics that differentiate the two categories (see Table 2—note the additional data point concerning support for English). It should be noted that some organizations involved in global ecommerce will not opt to have a multilingual website. The common alternative is to have localized monolingual websites. Lands End, for example, has its headquarters site in English and has subsequently added two monolingual, localized sites in Japan and Germany.
Base = business establishments with Web access whose organizations currently have a website
Base = business establishments with Web access whose organizations currently have a website
Regional variationThe dominant language supported on a website varies by geographical region (see Table 1). In North America and Europe, the dominant language is English. However, in Europe, German and French also stand out. Table 1 excludes data collected for several, less prominent Western European languages. In Asia/Pacific, the dominant language is Japanese, followed by English. In Latin America, it is Spanish and Brazilian Portuguese. Although English is never ranked lower than third, it is the majority language only in North America. Table 2 and Figure 1 show the role that English plays, either exclusively or with another language, on the websites of those respondents interviewed in the eWorld 2001 Survey. English is used on the majority of sites only in North America. It is also interesting to note that Asia has the highest proportion of multilingual sites. Figure 1. Worldwide website Language Support by Region. 
The Importance of LanguageEarlier Project Atlas II data (see Language Preferences on the Web, IDC #21181, December 1999), republished here in Table 3, indicates that most non-English speakers (as many as four out of five respondents in countries such as Germany, Japan, and China) prefer to use their own language on the Web. The point is that language counts. Although English holds a dominant position as a "second language," IDC predicts that the naïve assumption that English is the language of global commerce will have less and less credibility as time passes and the demographics of the Web shift away from English and the United States.
Notes:
Size of SitesTable 4 shows language support by site size (small, midsize, and large). Large organizations are the most multilingual (41.2% of large sites), and small sites are the least multilingual (22.1%). Stated another way, multilingual support increases with organization size. Three out of four sites support English, but the highest percentage of these sites are owned by midsize companies (82.1%).
Base = business establishments with Web access whose organizations currently have a website
Industry SitesWorldwideThe worldwide sample was analyzed by five broad industry categories, including an "other" category (see Table 5). The Manufacturing industry had the highest percentage of multilingual websites (34.6%), followed closely by the transportation/telecommunications/media/utilities industry group and the "other" group. Although the differences are not great, they are statistically significant. The financial/banking industry is the least multilingual and also the one most centered on English.
[*Transporation = Transporation/Telecommunications/Media/Utilities]
United States and CanadaAlthough the United States and Canada have the lowest proportion of multilingual websites of any geographical region, it is still interesting to see the industry lineup. The stratification of the U.S. and Canadian versions of the eWorld 2001 Survey accounted for more industry categories than those of the other geographic regions. As a result, we can see four times as much detail (i.e., 20 industry groups instead of 5). Each industry group had at least 50 respondents. Even so, to keep the resulting analysis statistically significant, we have collapsed those 20 groups into 10 (see Table 6).
[MnL = mono-lingual; ML = multilingual; 1L-E = one language (English); 1L-NE = one language (non-English); ML-E = more than one language, including English; ML-NE = more than one language, excluding English, UVN = Unweighted Valid N]
Professional services, driven primarily by IT services, is the industry group with by far the highest percentage of multilingual websites (42.8%). The next tier of industry groups (in the 21–24% range) includes retail/wholesale, healthcare, and education, in that order. At the bottom of the list are banking, insurance, and "other" financial services. As with the worldwide sample, these last two industry groups are the ones mostly heavily focused on English. The only multilingual sites that excluded English were in the manufacturing sector. Target AudienceWebsites that solely target other businesses tend to be multilingual (see Figure 2). About 56.3% of sites were multilingual if the intended target audience was other businesses, whereas about 64% of sites were monolingual if the intended audience was either consumers only or both consumers and businesses. Figure 2. Worldwide Website Language Support by Target Audience. 
N = 2,996
IDC believes this reflects the greater importance of business-to-business (B2B) ecommerce over business-to-consumer (B2C) ecommerce. IDC's Internet Commerce Market Model shows that B2B commerce, measured in dollars, accounted for 78% of all ecommerce in 2000, and that percentage will rise to 85% by the end of 2004. In other words, B2B is where the action is. It should be no surprise that as ecommerce encompasses a greater proportion of the generally expanding global economy, emphasis is being placed on accommodating the languages of international traders. This requirement has driven major B2B software vendors like Ariba and Commerce One to add local language support to the software applications that are used by international traders. However, enterprises catering to consumers frequently only cater to local audiences. These companies have the option to create monolingual, localized in-country versions of their websites (e.g., Yahoo France), instead of multilingual sites. They will be more likely to succeed if they can service prospective buyers in their native language, as indicated in Table 3. Hence, consumer sites are more likely to be monolingual, localized in-country sites. Site ActivitiesOrganizations with multilingual websites are more likely to have customer service that is integrated with a call center (see Table 7). They are also likely to have support for online auctions, another business-oriented activity. In the same vein, multilingual websites are more likely to have self-service customer support. On the flip side, organizations with monolingual websites are more likely to have community features such as bulletin boards, free services such as email and personal Web pages, and detailed information about the organization's products and services.
N = 9,638
Efforts at LocalizationTable 8 shows what companies in different regions are doing to localize their websites. If we look at the regional variations for each category, we can see which region does the least and most. The majority of South African sites (53%) provide content for other countries, but only a small percentage (14.9%) of North American sites do so. The percentage of sites providing price lists in foreign currencies ranges regionally from 53.4% for South Africa to 34.6% for Europe. The range for sites accepting foreign currency show a wider spread from 58.5% in Latin America to 21.5% in North America. International mirroring seems to be practiced most in Latin America and Western Europe. Regional warehousing is practiced most in Latin America and Asia/Pacific. The provision of local or regional service representatives is the most widely practiced localization activity, with at least three regions reporting more than 50% of sites using this practice (60% in North America).
Base = business establishments with Web access whose organizations currently have a website
Figure 3 shows that the sites providing content for other countries or regions tend to be the multilingual sites (no surprise), but they are limited to those sites that include English as one of the supported languages. 
N = 9,642
Primary Person Responsible for Site's Internet StrategyOur U.S. and Canadian respondents were asked who was primarily responsible for the policy with regard to the Internet strategy of their organization. Table 9 shows the responses for monolingual and multilingual sites. If we ignore the "other" category, the important contributors are top management and marketing and IT managers. It is not surprising to see that the marketing manager plays a greater role at the sites that are multilingual. Most sites exist to service customers and prospects; hence, whether or not to have multilingual support would inherently be a marketing decision. It is also interesting that the responsibility is more widely disbursed among multilingual sites.
Base = business establishments with Web access
ConclusionMultilingual websites can be distinguished from monolingual sites. Some of the characteristics that differentiate multilingual sites from their monolingual counterparts are site size, industry, and geographical region. Differences also exist in how Internet policy is set and in budget allocations for Internet initiatives, as well as the intended target audience, the mix of activities at the site, and the steps being taken to localized the website. By knowing these characteristics, vendors of globalization support software should be able to focus their marketing and sales efforts. (smcclure@idc.com) is IDC's Research VP for Collaborative Technologies, Content Management and Retrieval Software, and Knowledge Management Software. His the primary research focus is on natural language processing, multilingual content management, globalization and localization. Previously at IDC he directed Internet Tools which focuses on all aspects of application development for the Internet, Intranet and Extranet. Prior to this he performed similar functions as the director of IDC's Object Tools service. Object Tools focused on all aspects of object technology, including object-oriented analysis and design, object-oriented programming; and object database management systems. Projects in this role included two ground breaking Object Technology multi-client studies analyzing all aspects of object technology in MIS shops and selected ISVs. The use of object and related technologies has also been addressed in Mr. McClure's survey of Advanced Technology Groups in U.S. Corporations. Mr. McClure also tracks other emerging software technologies. |
|
LISA Business Data Forum Summaries and Presentations LISA Globalization Consulting Network Webinars and TouchPoint Advisory Calls 
LISA Forum USA LISA@Chinasoft Fair LISA Forum Asia LISA Forum Europe LISA Forum India Open Standards • TBX • TMX |
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||