Our Analysts' Insights

The history of standards for data and file exchange formats in the language industry goes back to the Localization Industry Standards Association (LISA) in the 1990s, which spearheaded the efforts around TBX, TMX, and GMX.  The Organization for the Advancement of Structured Information Standards (OASIS) organized the DITA, ebXML, XLIFF, and many other business data exchange standards. Linport is yet another initiative for localization data exchange. Most recently, GALA has been coordinating a new push called Translation API Cases & Classes (TAPICC). This latest effort is currently a pre-standards initiative, with participation from a range of industry contributors, to build industry-consensus agreements on metadata, use cases, best practices, and API classes.

Coordinating the interests and efforts of otherwise competing industry players and getting to meaningful results that enable continued innovation can be a challenge. In some cases, the core technology that underlies business application development changes faster than agreements can form around standards. Some question whether standards even matter, as they appear to be a lagging indicator. Why should industry participants care enough to put energy into developing them?

CSA Research sees the following reasons to get involved:

1. Untranslated assets impede customer experience, market share, and brand value. Translating information assets and localizing digital experiences result in better customer acquisition and retention, ultimately driving higher valuations for companies and brands. Yet global enterprises leave important content streams untranslated due to time, cost, and complexity of integration. Thus, enterprises should insist that vendor technology meets interoperability requirements, including service-oriented architecture (SOA) design.
    
2. Language service providers (LSPs) experience damaging delays when onboarding new clients. Though integration costs may seem “high” for customers, they hurt LSPs more than they help, even in cases where providers win the integration contract. This extra step delays the service provider from getting to the all-important recurring revenue that follows an integration. The industry loses revenue as well when customers decide not to proceed at all. A professional services organization may rely on customization to drive revenue for that team and manager, but immediate commencement of translation services through plug-and-play integration based on open standards would provide better results for the company as a whole. Time to revenue is an important factor in the cost of sales for LSPs. Standards minimize delays in both the sales cycle and the integration phase, ultimately reducing the cost of customer acquisition and onboarding for LSPs.
    
3. Technology vendors retard revenue growth due to inadequate standards adoption. Software suppliers are kept at the shallow end of the pool, unable to capture all the content that should be flowing through translation due to lack of standards in general and lack of adoption for those that do exist. No one vendor is at fault, as it takes cooperation from stakeholders throughout the technology value chain, from infrastructure providers to independent software vendors to service providers. Participating in or following industry initiatives is in the best interest of any developer of content and language solutions.
    
4. Standardization will happen around someone else’s needs. Industry players often expect to gain “free rider” benefits from standards, whether or not they help define them. The danger in this wait-and-see attitude is that standardization will happen around their competition’s needs and priorities, giving them an edge in implementation and time to revenue. Solutions succeed best when all of the relevant players come to the table and define truly common approaches rather than allowing a few parties to entrench their own approaches.

For continuous localization, the TAPICC initiative also seeks to support string-based content flows, in addition to file-based processing. String-based exchange allows systems to pass translation units or segments between tools and microservices without the overhead of files. A consensus on this topic would patch an important hole in the standards coverage to date. CSA Research recommends that enterprises, LSPs, and software firms alike should stay involved or informed as the TAPICC initiative progresses.