Many of us have spent years inside corporate America working hard to obtain the necessary budgets and resources to create a localization function so we can reach international markets. For those of us who have finally succeeded in making the case for localization, hiring the people to do it, and developing the necessary processes, it may be difficult to think of turning around and eliminating this function. Nevertheless, if we wish to truly globalize our businesses, that is exactly what we need to do. This, however, requires a vision that starts at the top and is understood and shared by everyone in the corporation.

Ten years ago, the software industry was in a very different situation, with most of the revenue coming from the U.S. As long as a company could remain profitable and sell only English products, there was no vital need to localize. As the balance shifted, and more and more potential was revealed in other parts of the world, American companies were forced to begin to localize their products. And that was usually the point at which they realized they had a lot to learn about developing products that could be taken apart for translation, put back together again, and still work.

Software developers were not the only ones who found themselves with a new level of complexity to their jobs; in fact, all across the organization, the implications of going global were felt. In areas such as technical publications, marketing, product support, training and sales, it was becoming evident that retrofitting was the hard way.

As technology continues to shrink the globe, more and more industries need to incorporate localization into their business strategies. Yet, despite the fact that many U.S. technology businesses now get at least half of their revenues from outside this country, Americans are still far from thinking, working, and developing product with a truly global vision.

For this reason, localization remains an expensive and often painful experience for both the client and the vendor, and one that still consistently requires super-human efforts on the part of employees on both sides to meet the requirements for today's markets. The turnaround times need to be greatly accelerated now, since products are updated much more frequently. There was a time when major international markets expected to get a localized U.S. product about six months after it was shipped into the U.S. and pay a hefty uplift for it. Now there is a reasonable expectation to get it 15–30 days after the product ships in the U.S., and the price differentials are dropping. What is more, we haven't really significantly changed the way we develop, test, market, sell and support our products, at least not enough to make localization a straightforward, predictable process. Before, we didn't dare start to localize until the product was "finished"; now we don't dare wait that long.

And yet localization is not rocket science—it's mostly a lot of hard work, attention to detail, and more than anything, real-time solutions for problems that should have been solved in the product conception and development stages. In many companies there is now a much greater awareness of the technical requirements for localization, but the knowledge is still neither complete nor widespread enough. Also, when U.S. ship dates come to push, it's the globalization requirements that usually get sacrificed. Sometimes the product is one that has been through many versions over many years and is still burdened with inappropriate legacy design and code. Sometimes it's a new product and is not yet destined for Europe, Asia, or Latin America, and so with the pressure to release in the U.S., globalization requirements are postponed.

However, the technical issues, although they have the most immediate effect on localization, are only one part of the problem. To be really successful internationally, an organization needs a strategy and an infrastructure to create and support an international business. Even if we finally do one day arrive at that magical moment when all developers have perfect knowledge of the technical requirements for all languages on earth—thus enabling parallel localization and simultaneous shipment of any number of languages—products cannot market, sell, and support themselves.

Executive attempts to determine the localization ROI (return on investment) formula that makes sense for a specific business are often devoid of context. The discussion focuses on the question "how profitable does a product have to be to justify the expense of localizing it?" Yet more pertinent questions are "In which markets do we want to compete?"; "In which regions are these markets viable?"; "What is the product set we need to succeed?" and "What time commitment are we willing to allow to determine the value of the return?"

A reasonable timeframe must be established to test the success of a product in a given market and region, as well as the specific criteria for determining success or failure, since it takes more than a fiscal quarter after introducing a product to establish a need for it. These factors then determine the investment the company plans to make and set the deadline for measuring profitability and re-evaluating the international strategy.

This is especially difficult for American software companies, since the very nature of the business defies long-term strategic thinking. The fact is that the commitment involves much more than simply localizing a product; it really must engage the entire organization to realize full potential.

In the early days of software localization, much was made of the cultural differences and requirements for products. There were many lengthy philosophical debates about language and images and the danger that American developers, in their insular ignorance, might risk grave offense to some other more ancient and sensitive society, whose members would then refuse to buy their product. Meanwhile, the reality was that it was already costing quite a lot to simply figure out how to translate the interface and documentation and put the product back together again so that it worked and not too many additional bugs had been inadvertently introduced.

If there was demand for the product in other locales, people were probably already struggling to use the English product, and would be happy to have the product in some facsimile of their own language, or at the very least to receive instructions on how to use the product in their own tongue. The economic realities soon revealed the need for a pragmatic and sometimes creative approach to providing a product that could be sold and used in different parts of the world, even if it did not perfectly match regional linguistic and cultural specifications.

As global communications shrink the world, and the free market system dominates international economics, we are finding that language purity is no longer a primary value. Even high-tech has to some degree created its own language, based on English as spoken in the U.S. (or at any rate an American programmer's version of English) since that is where the vast majority of the software products are created. Multimedia is a good example of this phenomenon: new words have been invented to describe functions and features that never existed before, and are universally understood by people working in the field. Internet communications is another example, where even the French have dropped the convention of using the formal form and now address each other in familiar terms. Machine translation, with all its hilarious imperfections, can be used in real-time on sites such at Yahoo for a quick communication between people who do not share a common language.

Nowadays, when companies need to become global to grow, and product information—including availability and pricing—is readily available to anyone with a computer and an Internet connection, customers everywhere are demanding the same quality and timing for product releases. This is influencing international organizations to concentrate on the similarities, rather than the differences between customers, and to focus on creating a product that suits the needs of the majority of its potential customers from its earliest conceptual phase.

Strategic global thinking also encourages creativity and appropriate economic compromises when striving to meet the needs of multiple international markets. Examples of pragmatic solutions to reducing costs are:

• creating one localized version only of languages found in multiple regions, such as Spanish, Portuguese, French and German,
• considering partial localizations when acceptable to the market, and
• being selective about which products to localize.

From a technological standpoint, the idea of creating a "world-ready" product is not new. Some of the more sophisticated global businesses have made significant progress towards this goal. The larger the organization and the deeper the pockets, the better the chances of setting long-term objectives that involve considerable resources. Establishing a strong technical foundation for developing products that can be easily and simultaneously localized is the fulcrum of an international strategy. But it is not enough by itself to guarantee international success.

Huge savings in localization costs can be realized once a product has been technically specified and developed in a way that makes it a straightforward process to translate the interface and help and reassemble the product. However, it is also important to include bilingual customers from different regions in usability testing. If there are any anomalies in the product's design that are essentially non-intuitive in certain cultures, apart from language issues, they can be uncovered and resolved at this stage.

The development of global documentation requires an international consciousness about the use of language, as well as the design and graphic elements. While it is not necessary nor recommended to write documentation for the purpose of machine translation, which is a technology that is not yet mature, it is always good to keep the writing direct, active, clear, and clean. Vocabulary does not have to be limited if it is chosen precisely and correctly, and documentation that native speakers can read and use easily is the very best documentation to be localized into other languages. It can be organized in a way that allows for partial localizations, when appropriate, and manuals can be designed as valuable single-source information that can be adapted for other purposes, such as help, training materials, and supplementary press books.

Global branding is vital for international success, both for companies and their products. Clearly the most successful international brands throughout history have been consistent all over the world, and effective in all countries. Presumably these brands were developed by marketing people from many different regions working together to conceive and test concepts and images. When successful, the result is the creation of powerful, universal symbols that are recognized everywhere and represent the same values everywhere they appear.

Marketing strategies need to be global as well, and should evolve early and simultaneously with the product's development. With the ease of communication in today's virtual organizations, there is no longer any excuse not to involve all appropriate people from the various regions in the design and approval of marketing campaigns. Naturally this can be challenging and politically delicate, but the extra time and energy required up-front pay off in process efficiencies and cost effectiveness in the end, especially when it comes to localization.

There should be a base marketing effort that is designed with the world in mind, and can be simply translated and used wherever the product will be sold. However, this requires the country of origin to show sensitivity to the elements that are truly unacceptable in certain cultures. This requires a careful and early review of concepts, since few marketing people are aware of all the cultural taboos that exist in the world. Nevertheless, there are fewer obstacles than is commonly thought to be the case, and by adhering to certain basic principles and avoiding a few specific images and phrases, this is not difficult to achieve.

The marketing people in the regions outside the country of origin must also be reasonable in their demands, and be prepared to accept materials that are not precisely what they would have created, but can serve their needs. The messages, images, and colors should be used consistently. A basic set of materials carefully prepared, reviewed, and approved for international use can then be localized very quickly and relatively inexpensively by using the same formatting, colors, and images, thus requiring only the translation of the words. If regions feel a need to supplement this base with additional materials designed specifically for their markets, they should have the freedom to do so, provided they can justify the expense in their own budgets, and manage the process themselves.

An international infrastructure is necessary to sell and support products in different regions. If a business decides to spend the money to localize a product and make the effort to improve the internal development process to support the localization, it must then provide the necessary resources in the local countries close to the customers. Well-trained salespeople, marketing people who are native to the locale, and knowledgeable support people are essential to assuring the success of a product. Often we see a company spend vast sums on the localization of a product that is not properly globalized, and then throw it at a region with no support whatsoever and wonder why it doesn't sell. It is then easy to say that localization is too expensive, and it's not worth the effort to continue pursuing those markets. This is especially true in emerging markets, where, for numerous reasons, even if some minimal support is provided, usually two or three versions of the product must be localized and sold there before it is possible to judge its success or failure.

If we envisage a company that has become truly global, and in which all functions are thinking internationally at all times, developing, marketing, and selling with the whole world in mind, can we then eliminate the localization function? To some extent, yes. By the time an organization has become aware of the importance of localization and competent at implementing globalization requirements, it usually has an established localization group. This group has perhaps been through a long, frustrating period of working to educate the entire organization and fighting to have enough resources to assure timely localizations in the face of many difficulties. As these obstacles diminish, the localization function should change.

It is usually at this point that there are many discussions about the appropriate localization model, centering around the issue of outsourcing. What can be outsourced, and what must be kept in-house? What resources need to be in a centralized localization group and what functions could be in the product groups? For most organizations, the outsourcing of localization is virtually a foregone conclusion. The question is to what extent can products be localized by vendors, come in on time and be of high quality? This, of course, is another large topic that would require another entire article. Suffice it to say that it should be theoretically possible to reach the point where the company's strategy is so global in nature, and the products developed so thoughtfully for an international market, that it becomes much easier to outsource the localization process. Under these circumstances, vendors do not need to have the intimate product and company knowledge that those inside the corporation have.

Fewer resources should be necessary within the company at this point, and these can be either central or distributed, or a combination of both, depending on what works best with the particular structure of the organization. However, if the integration of an international perspective has become complete, the people who are exclusively dedicated to localization should be few. Once development engineers can build localization kits, QA can do final validation on localized products, technical publications can support the vendors and validate the submissions, product management can specify the international requirements for the product, and marketing can create strategies that work everywhere. Then the sales people had better sell like crazy, no matter where they are.