My first IMUG event was beyond expectations! Fantastic insight, very smart questions, and incredible comfort of international and generally diverse audience. I signed up for a year 😉
The event started with a definition of a startup. Tex Texin, Chief Globalization Architect, XenCraft, noted that even if many opinions exists, the startup is a company working on something that has not been done before. A small business entering established industry with a traditional business plan would be a small business rather than a startup. Disruptive technology is the essence of startup, and internationalization for this technology could also be unknown (for example, gesture-driven interactions with a screen – how different the natural gestures would be in different cultures?).
in a startup environment everybody has to understand internationalization – there is no dedicated department and no time to correct business process or product features that have not been done with internationalization in mind.
Igor Afanasyev, Director of Localization, Evernote, described his company’s localization journey and probably received the most of the questions from the audience. This journey was the process a startup would need to navigate if it desires to expand beyond its home country. Interesting, 70% of Evernote customers come from outside of US.
Evernote had to find a way to automate localization. It was Russian/US company, and localization in Russian was the first step to establish the process that can be used in the future. No tools were available on the market in 2008, when Evernote started its process – it had to create its own tools. It took 7 months to localize Evernote for Russian market.
Russia was a test market, and its selection was based on company’s history rather than specific business need. After the process was established, Evernote identified profitable markets and eventually expanded localization to over 20 languages. The language selection was based on user data – Japan adopted the product even in English version and was a good market, so it was quickly included into the localization plan.
Company tailors content more for Japan and China and sometimes provides different content for the web site.
At certain point, the company brought translators in-house to improve quality and to assure that translators understand company better. Another challenge – managing people – accompanied the improvement 🙂
Evernote translation split: 75% – marketing, 25% – software. The process is automated around Google Docs. The company is planning to open source its system sometime soon. Engineers do not touch localization resources, they know how to internationalize strings properly. The company started localization early and it became a part of the culture; there is no specific person responsible for the process. Some of the challenges came from early decisions, such as selection of fonts not supported in all languages.
Are Evernote translators using machine translation? Yes, to help the translation process; translators are able to work faster if they start with machine translation. However, if a translator used Google Translate and changed just one word – this person is paid for this one word and the review of the translation.
Evernote historically supported “all brand new platforms” as a strategy to access early adopters. Some of these platforms no longer exist.
Evernote believes developing its own localization automation tool significantly benefited the company. Website is automated as software; marketing materials are translated by “Google Docs” method. Designers are educated not to have text in images; sometimes, different images are created for different languages.
Sometimes, the translation agency is becoming a part of a well-established process that a company already has. Sometimes, however, a marketing manager would come to Gengo and confess: “my boss told me to translate emails to French…” This situation is more “interesting”, and Gengo in this situation need to establish a workable process.
There are different combinations of resources that companies can consider:
- in-house translators would be better educated about your business
- translation can be crowdsourced if cost and time are critical factors
- in some cases, Gengo translators might be doing the translation, but the internal translators act as editors, who check proper brand and messaging. This approach is used by companies with an internal team of translators, which lacks needed capacity.
Non-traditional tools in localization
Should the company buy a tool or build it? If the company’s core skills are not in globalization, it is better to buy a tool. From another side, the company can discover interesting approaches if it ventures into creation of its own tool.
Questions to ask:
- where are my company’s core skills?
- is my product that different that it requires unique localization approach?
- is there added value in the creation of the tool?
(I guess, the process will be similar to email and marketing automation tools adoption. At first, many companies attempted to build their own “free” tool, and eventually, IT departments realized that they were not in business of “building marketing software.” Eventually, the tools became very advanced and sophisticated; specific marketing knowledge needed to create the tools were beyond internal IT departments. Now it is unlikely any company will build it own system. The adoption of localization tools will probably be similar: as tools mature and become increasingly valuable, more and more companies will adopt them.)
When do you add another language? It is a combination of:
- business opportunity
- cost of development
“If you are a startup, you might be selling to enterprises, and enterprises will pull you into different markets…”
Excellent event! The event was held in the Box office (great location from my perspective 🙂 ). I have been using Box for a few years personally, so it was a pleasure to visit the office.