Finally! Data to Prove that Localization Matters

By Donald A. DePalma

One of the most common refrains that we hear from software vendors and from too many enterprise planners is that “their buyers and developers can handle English.” They tell us that as long as the products can ingest, manage, and publish information in other languages, that’s fine. However, “our CMS (or database or ERP) administrators and application developers can get along just fine with English, thank you.” How valid is this point of view?

Unlocalized Products Cost More to Buy, Maintain, and Enhance

At Common Sense Advisory, we have long held this position to be suspect. We talk a lot with companies around the world and find that many do employ lots of Anglophone staff in their IT departments and that English might be the “official” corporate language. However, when we probe deeper, we hear about the increased expense of running English-only products. They tend to pay more for bilingual developers and to provide support and training for English-language products. Furthermore, the availability of fully trained, competent developers for your favorite CMS might be lower in a given country, thus raising the total cost of ownership for that product. We find that this “English-only tax” rises as you go further south or east in Europe, or head off to Asia. What’s the solution? We always argue for buying localized software; that is, products whose interfaces and documentation have been translated, and whose features have been adapted to the business, cultural, or logistical needs of its target markets. In an ideal world, you’ll find that the supplier for your U.S. headquarters’ CMS also supports French, German, Japanese, and Russian variants, with the administrative interfaces and developer tools adapted for use in those markets. This capability lowers the cost of developing, deploying, and managing applications in your IT centers around the world. If all variants share the same data model and underlying technology, then this internationalized stack will let you deploy the right tools to the right people in the right countries. Sound familiar? That’s the goal of other corporate initiatives like customer relationship management, personalization, and website globalization. Why shouldn’t your administrators and developers benefit, too, from the same intention to provide tailored information and interfaces?

Testing the Hypothesis that Localization Matters

We decided to test our hypothesis more systematically, so we polled 351 businesspeople in eight non-Anglophone countries about the software they buy for their firms. One of our goals in undertaking this research was to see whether software vendors can increase their sales by localizing their products and websites. For more details on the questions we asked, our survey samples, and who responded, see “Localization Matters” (Nov08). We selected a cross-section of countries around the world, aiming for a representative mix of markets for which companies frequently localize their products (France, Germany, Japan, and Spain), attractive developing markets (China and Russia), and locales for which English is often thought to be sufficient for most offerings (Sweden).

The Need for Localization Spans the Entire Buying Cycle

We asked our respondents about eight factors in the buying cycle where localization plays a role. The initial decision to purchase a given product over another begins with marketing literature and technical specifications. Usability enters the equation, with both user interface and product documentation. Technical support rounds out the buying cycle for when things don’t quite work the way that the manual says they should. First, let’s look at marketing materials. Items such as printed brochures, website descriptions, Flash animations, and testimonials grab the interest of potential buyers. Then they move the prospects from consideration to purchase, helping them understand a product’s value proposition. We asked our survey respondents to agree or disagree with the statement, “Having printed marketing and other collateral material in my language makes my organization more likely to purchase a software product.” Across our entire sample, more than 80 percent told us they agree with that statement. As we expected, the outlier here was Sweden, with just 60 percent of respondents preferring marketing materials in their own language. If marketers accept the conventional wisdom that English is enough for Sweden, they still leave three out of five buyers on the outside looking in. That’s not good for long-term sales. The other seven countries all came in at 80 percent or higher. We also asked about the desirability of adapting the product interfaces it to the linguistic, business, logistical, and cultural nuances of the country in which it is sold. This process involves a range of system components:

Data to Prove that Localization Matters 3 Copyright ? 2009 by Common Sense Advisory, Inc. 26 March 2009 Unauthorized Reproduction & Distribution Prohibited www.commonsenseadvisory.com

              ? End-user interface. About 85 percent of buyers see the value of having a localized user interface. Without access to all the capabilities of a product, they might use fewer options, rely more on defaults, and experiment less with product features. Respondents from Brazil, China, and Japan led the charge for localized interfaces.

              ? Server and web-hosted software. As companies expand eastward and southward into Europe and Asia, more than 80 percent of our respondents think that both sides of the network connection should speak the local language. Respondents from China and Brazil are more likely to buy client/server or web-hosted software when it has been localized to national practices. Russians and Swedes were less committed to localized goods, but the majority in both countries preferred localized server products.

              ? Application development. Nearly 80 percent of our respondents prefer to buy programming tools that have been adapted to local practices. Localized software allows a wider range of developers – including those not conversant in foreign languages – to create value-added applications for their companies. This capability is especially important for an enterprise product meant to be integrated with other applications such as database, content management, and enterprise resource planning. Buyers from China prefer localized tools the most, while Swedes are the most English-tolerant.

Proficiency Influences Purchase Decisions

Our respondents told us that having information in their languages, from first contact to post-sales support, made them more likely to purchase a product. As a final test for localization preferences, we aggregated the proficiency data for end users, administrators, and developers into a single number. Then we analyzed the buyer responses to the questions of whether these various constituencies could handle English-language products, yielding the percentage for that cohort in Table 1. Depending on the country, our respondents feel that 33 to 69 percent of this user cohort lack the language skills required to use English-language software. Then we combined the various components that make up a product offering – translated information, localized software, and in-language technical support – to create an index of purchasing “desirability” from a localization perspective. The third column of Table 1 lists the percentage of buyers who, when considering all the descriptive, functional, and supporting elements of a complex product like software, told us they would be more likely to buy that product if localized. The numbers range from four out of five in Sweden to nearly everyone in Germany, Russia, and Spain to all buyers in Brazil, China, France, and Japan.