Japanese crowd translation startup Gengo claims they sold around 3.8m words last quarter. At a word price of USD 0.06 that’s between 230.000 and 304.000 of quarterly sales. They pay about half of that to translators (word price around USD 0.03).
So their monthly sales after purchasing expenses are around USD 50,000 (USD 600,000 p.a.). From that they still have to pay staff, etc., so they’re probably not enormously profitable.
Now they might have long-term contracts, which guarantee additional income, but I don’t see why a TL buyer would do that, considering there are plenty of agencies selling the same product. I haven’t seen them announce any life-saving business partnerships.
Google’s partner for “crowdsourced” translation is Translated.net (Italy; same shitty price, shittier technology, but hey, they’re partners of Google). AFAIK, Gengo mostly sells to fragmented online retailers, who in turn all have to compete against “ebayZon”.
Google itself invests little in crowd translation; their Translation Toolkit and their machine translation haven’t evolved much in recent years, so they prefer to buy instead of make this service.
Now Gengo just got USD 5.4M in round C funding. They do have an “interesting model”, i.e. they’re smart skimming our wages, but I don’t see how they want to get more out of a business.
Maybe once they go public, Gengo can start paying their translators in stocks; they won’t be worth more than their avg. word price.