1Mby1M Deal Radar 2010: Lingotek, Draper, Utah

Even as companies large and small become more global in ambition and scope, language barriers remain a fact of doing business across borders. Language translation is one of those costs that most people don’t automatically think of when they’re looking at a balance sheet. But for big organizations, translating hundreds of thousands of words into another language can be a significant expense. The average translator charges 21 cents a word. This paragraph alone would cost $17.85 to translate. Lingotek makes software that aims to disrupt the traditional translation model. The company has developed an Internet-based translation platform to give clients more control over the translation process while cutting the average cost of translation by more than half.

The Draper, Utah–based company was founded in 2005. As a child, president and CEO Rob Vandenberg lived abroad with his parents, a journey that included a stint in Saudi Arabia. Language – and overcoming language barriers – has always been a big part of his life. He graduated from UC Berkeley with a degree in political economics and has worked in B2B tech ever since. He was one of the first employees at INTERSHOP Communications and went on to launch LocalVoice.com as that company’s cofounder and CEO. LocalVoice was bought by HarrisConnect in 2005. He worked there as VP of sales and marketing before joining Lingotek. He’s been the CEO of Lingotek since 2008.

When Lingotek entered the market, language translation was being done essentially the same way it had been done for decades: Companies would send their documents to an expert, and they’d pay them an average of 21 cents a word and get their document back in another language. This was before Google Translate or Microsoft’s machine translation tools. There weren’t, says Lingotek, tools out there to help companies cut out the middleman.

Lingotek offered an online translation workbench that integrated machine translation and “translation memories” – intelligent dictionaries that are capable of learning company-specific language. It claims to be the first company to introduce crowdsourcing and community voting to the translation process. Its platform comes in hosted (subscription; $150 a month) and on-premise (licensed) versions. Companies upload documents to Lingotek’s Collaborative Translation Platform, which scans those documents for common phrases or sentences that the company has successfully translated in the past. The platform is capable of learning company-specific terms, making the translation process faster as time goes on. The untouched portions of those documents are then ready for human translation. Lingotek’s document management system is designed such that multiple translators from around the world can work on a project simultaneously and even give feedback on the work their fellow translators are producing. The company can also embed any feature of its application through a set of APIs that allows clients to translate content inside any Web application such as SharePoint, Drupal, Salesforce, Jive Social CRM, Oracle UCM, and more. It also runs an in-house translation service to help companies that might not already have a base of translators on board. It remains a bit unclear, however, how professional translators benefit. In the recommended reading on machine translation, Lingotek’s former CEO puts forth the argument that translators can become more efficient with Lingotek and increase their earnings even as their clients save money. But in this 2009 presentation, Lingotek presents a model in which companies use crowdsourced translations where translators, who may or may not be professionals and may or may not be working on a volunteer basis, are paid fractions of a cent per word or not paid at all.

The worldwide translation market is worth roughly $15 billion a year and is growing at 8% year-over-year. Lingotek believes its collaborative translation platform is capable of disrupting this market and taking a sizable chunk of it. But as intractable a problem as machine translation has proven to be, there are several other companies in addition to Microsoft and Google that offer translation software, including NativeTung, CelloTrip (which focuses on short messages), and Dubzer. At present, the company is targeting Fortune 500 companies that have presence in global markets and require smarter language translation software. Its customers are mostly mid-market companies with at least $200 million in revenue. Its first major enterprise win was a translation contract with Adobe Systems, a contact it made through the company’s director of globalization, Francis Tsang. Vanderberg met Tsang several years ago at conference in Belfast, Ireland. He was working in a similar capacity at Apple and Oracle, leading the charge toward greater efficiencies in globalization.

Several hundred companies use Lingotek’s software, including government organizations and large enterprises. Enterprise customers include Adobe, Novell, eBay, and Avaya, among others. For one particularly large project, Lingotek helped the Library of Congress translate nearly one million words into Arabic, Chinese, French, Portuguese, Russian, and Spanish for the World Digital Library.

The company has raised a total of $5 million in two stages, most recently in March from In-Q-Tel. Its lead investors are Flywheel Ventures and Canopy Ventures. Both Flywheel and Canopy are leading VC firms based in the Rocky Mountains, focused on emerging technology coming out of that region. They were attracted to Lingtotek because they felt it had a large market opportunity and an unfair competitive advantage in the space. They brought on In-Q-Tel, which is best known as the CIA’s investment arm. In-Q-Tel is chiefly interested in technology that can benefit the government, and Lingotek says that the company has introduced it to a lot of interesting government contracts. It’s not looking for more funding at this time.

The company is on track to bring in $2 million to $4 million in revenue this year, and its goal is to be profitable by Q4. It’s doubled in size every year for the past three years. Lingotek is focusing on three growth areas: federal sales, commercial sales, and translation services. Its federal business is self-explanatory: software sales and licensing to various branches of the government that employ teams of translators and need a central platform to manage their work. It’s similar on the civilian side, as the company continues to recruit Fortune 500 companies interested in speeding up their translation process, regaining control of the workflow, and cutting costs.

Lingotek’s ultimate goal is to be acquired and believes that it has several characteristics that make its business attractive to an outside buyer, including its technological advantage, strong relationships among federal and enterprise customers, and soon, its profitable revenue flow.

Recommended Readings
Lingotek to help US defense agencies share data (from TranslationAutomation.com)
Is translation memory obsolete? (An older but thorough discussion of machine translation)

This segment is a part in the series : 1Mby1M Deal Radar 2010