It has been rather busy at our office during last couple of weeks. Additionally we’ve been busy meeting interesting people at different. Sadly we missed #Pyconfi at Turku last week but I hear it was a great event. We just love Python and Django (that powers Get Localization) and it would have been interesting to see what other people have created using them.
But we went to Slush10 and we are happy to see that localization is something that startups really get. We talked to many people and saw many good presentations. One of the things that got our attention was that many startups mention localization as a necessary step in growth of a service and reaching new markets. We couldn’t agree more on this. Localization is not just something you do after you have grown. For many services the localization is the means for growth (see the story about Runtastic). Web services and especially mobile services are really personal. When you serve people personally you serve them with their own language. Nokia’s EVP Niklas Savander put this very nicely in an interview yesterday: “Location not just navigation will be big in many markers. ‘we have the availability of locally relevant apps. It doesn’t matter if you have 200,000 applications in Vietnam if they are all in English.'” See the story here.
On other news you may have noticed a story by Talouselämä later on picked up by Arctic Startup on how Nokia is putting more emphasis on crowdsourcing and they have a chief of crowdsourcing. Story by ArcticStartup here and original Talouselämä arctic in Finnish. Article gives you a good insight on how to plan and organize your crowdsourcing project.
Translating your iPhone application is becoming more important day by day. US is still the biggest market but others are growing fast. Based on AdMob report, iPhone is growing in Europe and Asia faster than in US. In Japan alone, growth has been 350% from January to November 2009.
Currently iPhone is available in 96 countries. Looking the pie on left, about 65% of AdMob (Note! These numbers are based on web usage, not the actual sales) traffic is coming from English speaking countries. Rest 35% is heavily fragmented and basically covers rest of the world. From this you could probably figure out already that you should localize your app at least to French, Germany and possibly Japan. You may even think that it does not make sense to localize at all as those three countries will cover only 13% of the cake. And I would say that you might be right if you think about localization like you do now.
Did you notice those others? Others is the second biggest part of this pie with 17% share. It doesn’t make sense to localize your application to these “others” as it means tens of different languages. It would cost you quite a lot in case you go with that route. Well adding up those 17% with other non-english countries, you end up to 35%. Nice number and in actual devices it means 27 million devices (total 78m). In potential customers it is 27 million potential customers more for your application. Our solution to make this happen with minimal effort and costs is crowdsourcing. It’s really difficult and expensive to handle all of those languages as the market is so highly fragmented with any traditional methods.
Basically by crowdsourcing we mean outsourcing the localization work to your users. With our platform, it is possible just to upload your original English content to our web service and give the link to your users. They will take care of the rest. You just download ready-made translations and compile them into your application or use our API’s to fetch them over the air. Let your users know about the possibility and they will translate if they like your product. Even if the product is commercial, it works (we’ve done it, see the previous posts).
Our goal is to take care of the whole localization process so that we help your users to do translations like they were professionals. Trust your users, they’re using your product which means they trust you. They know how your product works as they are the actual users. Because of that, your community can take care of the translation, validation and even localization testing on your behalf. This also means that you don’t have to guess the languages you translate, your community will translate the languages they need. This is how you turn your translation projects to lean projects.
We strongly believe that localization is one key element of successful application or service. Good example is Facebook which basically skyrocketed after providing their service for other languages as well.
I’ve personally crowdsourced successfully my application localizations quite long time already in my hobby projects and saw that for some people, it’s extremely important to have version available in their own language. After turning these hobby projects to business, supporting multiple languages started to feel difficult and I decided to release commercial version in English only. That was due to the urgent need to turn the application fast to commercial solution. Managing multiple languages, starting from application level support to translations and testing is really time and money consuming process so for small company like us it’s not often even a good solution. It could go once but there’s always next release around the corner so how in earth you are going to support multiple languages in this agile and lean world where you’re suppose to release often, release early.
Well at least that was the situation yesterday. After recognizing this problem, we’ve been working with the solution (actually already from spring ’09). After all, by localizing we can grow our potential market and create more revenues so it’s worth it. We had three requirements for localization system:
It needs to significantly decrease the overhead needed by localization.
It needs to natively support the platforms we develop on (specific localization files etc.)
We noticed that there’s no tools for our needs so we started to develop our own. The solution consists of two simple elements:
Location Management System (LMS). Something similar for localization what we have for source code already e.g. SCM.
Crowdsourcing: our users are going to translate our applications. And yes, they are willing to help us. Why? I’ll give an answer in upcoming posts.
This allows us to create separate process for localization which is running concurrently with our development processes. Which is exactly what we need, no more hassle with the strings. I just post the modified or new strings and they get translated.