WinterGreen Research came out with a 668-page report last month about the machine translation market. This report estimates the machine translation market at $2 billion in 2009, growing to $7 billion by 2015. These numbers are strikingly different from Language Weavers $67.5 billion estimate of the "untapped" machine translation market by 2011. Why is that? Well, just as has been noted about the general translation market, the size of the machine translation market depends largely on how it is defined. In this case, WinterGreen Research does not appear to include "untapped" potential, but in these numbers it does broadly include translation services and non-MT tools that may be used in conjunction with MT. When asked to clarify its definition of the market, WinterGreen Research said,
The report profiles the machine translation market, the services and maintenance market, the language translation services directly associated with the machine translation. The machine translation is used in many industry segments and there are discussions of those different segments and how the software is able to support efforts to achieve localization in different regional and industrial market segments. This study is a discussion of the translation tools market to the extent that the tools are part of machine language software and repositories used to facilitate accurate translation by humans....
I have not commented on this report until now because I needed to obtain some clarification regarding the topic of the report. The description of the report has a heavy SCIgen-like feel that makes it difficult to determine which market is actually being discussed. Let me know if you encountered the same challenge. Read the description here and please tell me what you glean. At first glance, does this report appear to cover machine translation, translation tools (i.e. TEnTs, TMs, CAT, etc), software localization, all human translation services, or a specific combination of these markets? Or does the report not distinguish one of these markets from another?
The title and the last two paragraphs really left me puzzled at first:
Language Translation Software and Services Market Shares, Strategies, and Forecasts, Worldwide, 2009 to 2015
....Worldwide language translation services software license shipments market share analysis indicates that LionBridge has 16% share of a $2 billion language translation services market. LionBridge and SDL are market leaders with IBM participating in the services portion of the market supported by its WebSphere product that offers portal technology...
Localization translation services based on software at $2 billion in 2009 are forecast to become $7 billion markets by the end of the forecast period. Software is becoming much more accurate as it combines the separate rules engine, translation memory, and statistical technique algorithms that have been used separately hitherto to support translation services. The combination of technologies is anticipated to create systems that are more accurate. These markets are part of a $10 billion larger translation market, forecast to reach $21 billion by 2015.
Contradicting my initial interpretation of the report description, the list of companies mentioned includes mostly MT providers and the table of contents leans heavily toward MT.
Fortunately, that has all been clarified now and you can read the report with the expectation that you will be reading an analysis of a broadly defined machine translation market. The entire report is likely even more interesting than the summary description. If any of you pay the $5,000-$10,000 for the report, please share your opinion of it and let others know if you think it was worth the investment.
Those who enjoyed this article may be interested in reading more about SCIgen.





