01
May
Benjamin B. Sargent and Donald A. DePalma 1 May 2007
Filed under (Translation Technologies)
3 pepper rating

For the last few years we’ve been arguing that translation memory (TM) should go the path of productivity software in the early 1990s — US$99 per translator or a low-cost subscription fee. Nobody bit on the cheap license, but Lingotek did offer any translator its TM tool for US$300 per year, while Lionbridge gave freelancers it employed free access to its Freeway translation environment. Lingotek got closer to our ideal price point but didn’t arrive, while Lionbridge got the price right but remained closed to anyone but its own translators. Meanwhile, SDL and other desktop stalwarts continued their time-honored big TM pricing with annual pricey upgrades.

Today Lingotek announced that it would follow our advice by dropping its individual subscription fee to zero. Earlier this month CEO Tim Hunt told us that Lingotek would make its money by installing its TM software behind corporate and government agency firewalls: "We can make money on private server licensing. That means we can let everyone use if for free on the web — and skip the US$300 per year per user fee."

For those who don’t know the company, Lingotek offers a pure web-hosted application for translation that is as much search engine as it is a TM workbench. It hosts many customers (that is, it’s multi-tenant) but allows TM sharing across users (cross-tenant). Most multi-tenant systems like Salesforce don’t support sharing between accounts — for example, you can’t share selected leads among partners. With Lingotek, if two companies want to share specific data or documents, they can do that — without involving a broker or high-level agreements among dozens of companies.

So what will "free" do to the market? Without a brilliant set of alliances, we believe the value proposition — that is, massive databases in lots of languages at no cost to freelancers, high costs with high potential return to corporate users with private versions of Lingotek — will develop too slowly to matter. How can Lingotek achieve YouTube-style rapid adoption?

  • Get under the skin of translators, especially freelancers. A free, highly functional computer-aided translation (CAT) tool will appeal to the 85-90% of translators not using TM, to accidental translators pressed into ad hoc service within their organizations, and to companies where the traditional uses of TM to translate documentation and localize software don’t present themselves.
  • Buddy up with translator portals. Lingotek should team up with portals like Aquarius, Proz, or TranslatorsCafé to reach critical mass fast. As salesforce.com and Webex have done, Lingotek could benefit from providing self-paced online training for new users and hands-on demos and sessions at major translator gatherings around the world.
  • Join the relevant content ecosystems. Sure, its spellchecker supports Kashubian and Kinyarwanda, but Lingotek doesn’t know from enterprise-critical formats like FrameMaker. While Lingotek has begun building out translation management features, its integration capabilities such as content connectors will be available only in the private installations — starting at US$750,000. Informal mash-ups or formal application integration with CMS and TMS will be essential ingredients to this strategy.

If this is a bid to unseat SDL-Trados, the price could bring in those freelancers and low-margin users, but the lack of enterprise integration will limit its effect. The TM supplier with technology most like Lingotek’s is Lionbridge, which we said should open its Freeway solution to freelancers, for free, a year ago. Lionbridge already has the scale and enterprise connectivity that Lingotek must achieve, but it doesn’t have the right business model to become the dominant translation memory solution.

The bottom line: The TM market has definitively segmented into old-model vs. web-model. The pure web version promises to drive far more usage than the over-featured desktop-resident CAT tool of yore.

Share or tag this post on: