6 months ago
December 3, 2019
Key Takeaways from SlatorCon Amsterdam 2019
On November 28, 2019, SlatorCon Amsterdam 2019 brought together close to a hundred language industry leaders at the striking Andaz Hotel in the heart of culturally vibrant central Amsterdam. On offer was a packed agenda that included globalization and localization strategies, launching a product into a new market, and much more.
First on stage at SlatorCon Amsterdam 2019 was Florian Faes, Slator Co-founder and Managing Director. Faes spoke about current and emerging trends in the USD 23bn (and growing) language industry. He highlighted pockets of disruption in the high-growth verticals of media localization and gaming, both of which are being influenced by streaming.
Faes also discussed the current state of investments in the language industry. He singled out one big change to the nature of such investments compared to five years ago: Back then, it was all about uberizing the marketplace if you wanted to get funding; now, it is all about AI.
Discussing the competitive landscape for machine translation (MT), Faes said the market is still trying to figure out how to interact with MT technology and what exactly to do with it. He said there is an opportunity for tech companies to come up with a product that allows translators and linguists to interact with it in a better way than traditional language productivity tools.
Jimena Almendares of Intuit tackled the question of how a big, public company views international expansion. She detailed the process that the Intuit team went through to bring their QuickBooks product to market in Mexico inside of just three months.
She said partnerships are key when launching into a new market — all the more so when choosing to expand in a country like Mexico, which has a complex financial accounting system. From HR and legal to regulatory considerations, there is much more to consider aside from localization.
Language Data, M&A, and Localization Maturity
Acrolinx CEO and Founder Andrew Bredenkamp discussed the role of data in AI, saying that data labelling accounts for a large portion of the work that goes into building AI systems. One thing about the fast-evolving AI space, he joked: “By the time we’ve got to coffee, things will have changed.”
Bredenkamp said it is easy to assume that AI can do everything, but there are limits. He added, however, that human translation will become increasingly rare for any of the big languages. Higher value activities, on the other hand, such as transcreation, will be left to humans…for now.
Patrick Prokesch of i5invest provided SlatorCon participants with an in-depth look at some of the investment models prevalent in the language industry. According to Prokesch, the language industry saw USD 0.3bn in VC and PE investment in 2019.
He explained that PE investment (the dominant form of investment in the language industry) is focused on mature companies that will go on to do a buy-and-build. PE looks to generate a 2–3x return on the initial investment; whereas VC investors, who have a comparatively high-risk appetite, expect higher returns in the region of 10x.
Then it was over to Hartmut von Berg of LogMeIn for his take on developing localization maturity within a large organization. Having gone through a merger in 2017, LogMeIn found itself with two separate teams performing localization, reporting into different functions; technical writing and engineering.
Rather than consolidating the teams, they built a dedicated localization team, which von Berg heads. The team’s goal, to transition from being an internal translation service provider to becoming globalization consultants to the business. The transformation is based on the strategic pillars of process, purpose, and organizational development.
Enterprise Panel, Personalization, NMT
Three other localization buyers came on next: Àlvaro Villalvilla Merelo of Nike, Andrea Guisado Muñoz from Kayak, and Vinicius Britto of Bose. They were joined by Balázs Benedek of Easyling in an enterprise and e-commerce localization panel, which was moderated by Slator Co-founder Florian Faes.
Topics included how each panelist handled the complex activity of SEO localization, the value LSPs add, and the level of machine translation (MT) usage in their respective organizations. The consensus was that it is important to take a differentiated approach when thinking about which content types are suitable for MT, with at least one buyer suggesting to not use MT for marketing content.
Michal Antczak of PayPal showed that localization can have an impact on the international revenues of a business, adding that it is fundamental for localization teams to secure the buy-in of key stakeholders within the business as they, ultimately, hold the purse strings.
In Antczak’s view, hyperlocalization is the next leap in engaging users; and full hyperlocalization of a product could help boost revenues beyond what is achieved through localization.
Capping the day was Yaron Kaufman from language service provider (LSP) One Hour Translation. He said customers are motivated to think about NMT either because of a desire to reduce costs or because they have too much content. Specifically: How can I take advantage of the NMT revolution? How much can I save?
Although still imperfect, Kaufman said NMT is good enough and, on average, customers using machine translation can expect to save around 30% of translation costs.
SCADM19 Presentation Andrew (Acrolinx)
SCADM19 Presentation Jimena (Intuit)
SCADM19 Presentation Florian (Slator)
SCADM19 Presentation Hartmut (LogMeIn)
SCADM19 Presentation Patrick (i5Invest)
SCADM19 Presentation Yaron (One Hour Translation)
SCADM19 Presentation Michal (PayPal)