Keywords said Kantan generated annual revenues of EUR 0.8m in its most recent financial year ended March 31, 2019 and achieved an adjusted EBITDA margin of 22%. If the full consideration were to be paid, it would translate into a purchase price of 8.75x revenues and 39x adjusted EBITDA. This sets an interesting data point for the dozen or so pure-play machine translation players that operate a B2B managed service model; as opposed to, for example, DeepL’s model, which is geared more toward B2C.
KantanMT was founded in 2011, at a time when statistical machine translation was still the standard, and transitioned to neural machine translation models after 2016. The company’s client portfolio includes eBay, VistaPrint, the European Commission, and language service providers, according to the announcement. The company currently employs nine people and is based at Dublin City University’s ADAPT Centre in Ireland.
Slator 2021 Language Industry M&A and Funding Report
46 pages on language industry M&A and venture funding. Includes financial investments, mergers, acquisitions, and IPOs.
The same announcement said Keywords had used Kantan as a vendor before. The acquisition is interesting; gaming is a notoriously difficult space for machine translation given the content’s highly creative nature and fan vigilance over how games are localized.
According to Keywords CEO Andrew Day, “Video games are highly context specific, often being set in fantasy worlds, are highly immersive, and therefore require both high quality and highly adapted localisation, are multi branching and use real time, AI generated content.”
Keywords has been highly acquisitive since its 2013 IPO, and the Kantan acquisition is in line with its strategy of bolting on niche capabilities and technologies. Alongside the Kantan deal, Keywords also announced the acquisition of London-based Ichi, a creative and marketing services firm, for GBP 3.2m (USD 4.22m), and Canadian audio recording and casting services company Syllabes for CAD 0.475m (USD 0.36m).