How to quickly localize your app for 300 dollars
blog.discofingers.com
blog.discofingers.com
If your target audience is international but you can reasonably expect them to speak your language but to search in their native tongue, OK, this makes sense. But that's a large target for a small segment.
Localization is a process that involves your translator understanding your product, the terms it uses, the context in which it uses them, and how those terms can be transitioned into their target language. That's a far cry from keywords and descriptions, and it matters in many cases. Those cases never stop at $300.
It's often worth doing (this is totally dependent on your goals, though) but in the case that it's worth doing for you, it's also worth deciding whether you want to actually do it or whether just making sure people find your app is OK or whether you want them to be able to use it in their native language. Huge difference.
However, I'd definitely not recommend people translating their entire apps without localizing it's metadata first to check your apps potential in various markets.
For a lot of apps, including our music making app, it's meaningful to reach out wide even though the language within the app is english, as it's not a very text based app.
The result will give us a good overview over which languages that has the most potential, which we later on can consider to localize fully for.
If you do iterative and lean development, localizing using freelancers website will really slow down your release cycle. This multiplies your fixed costs you spent running your small company.
Another point, you'll also need to pay others skilled translators to verify the quality o the localization.
In my case, after localizing it, the numbers of user increase just a few. If I would spent this money on marketing and development, I'm sure that I could get better results.
It will make it cheaper and better.
Interesting website. Do you get to communicate directly with the translators?
It also looks like you can respond/open an issue, but I haven't tried that.
Really the most annoying part was updating my screenshots, hours spent in Photoshop. This should be much easier since I've started using Sketch though (Tip: convert screenshots to symbols, choose "Exclude Text Value From Symbol" for your text layers, so you can duplicate the symbol and edit the text many times.)
The increased downloads will allways be relative to the downloads you have in that spesific market before the localization.
It is true that you'll have to spend time doing it, which translates into money. However, I claim that you can do it using only three hours, plus the time it will take you to adjust the screenshots (it will normally be a design job in addition to just changing the text.)
These translators are good at their job. IMHO you won't need to double check their work with other translators.
"If you do iterative and lean development, localizing using freelancers website will really slow down your release cycle."
How come? Because it's freelancers you have reached through a website?
Localising like this might get you more downloads in a foreign market but you are potentially burning bridges with customers because you cannot give them the support you give your english speaking customers.
Sure, you may earn money from it, but it's not really helpful to your customers.
If a russian person searches for делать музыку (make music), why is it spam that our app is there?
Case in point: I am that "russian person". When I search in Russian, I am doing it because I want to get russian apps - that is, fully localized apps whose UI is natively Russian. I positively hate it that the app store is full of apps that were "search optimized" like you suggest.
Very interesting. Maybe anybody has figures like this on translated meta-data of 'ordinary' websites ? I mean not in particular for any downloads, but for the impact on the number of website visitors.
Google will most definitely crawl all your localised pages, so you'll at least get more local traffic.
Dont have any numbers though..
How did you vet the translators that you found? Like you said, there are lots of poor translators out there (and even professional translators make mistakes). Without having a second translator review the translation, how can you be so confident that no mistakes were made?
You can be 100% confident mistakes were made.
This method is useful for apps that don't care as much about quality.
Language skills: I made an Odesk job for each language and got quite a few applications. THe first filering process consists of deleting those who have below 4,5 stars, and thos who have't worked at Odesk for quite a while. Ideally 1000+ hours.
The I sent a sample to the most promising candidates. These samples were all sent to professional translator, who picked their winners.
Marketing skills: I asked every single translator how they would research keywords in their local language. Their answers will easily give you an overview of how reflected they are in this field. Plus, I only chose people who had done similar tasks before. Not everybody had done app localisations, but at least localisations for web marketing purposes.
I can't be 100% sure that there are NO errors in the translations. But I can be 100% sure that this localisation will benefit my business, as it will increase downloads and revenue significantly. I know this from previous experience.
long answer: Most Portuguese users under 30 years old can read and write, at least, basic english, and would prefer to use the english version of your app instead of brazilian portuguese. This is the scenario that I experience with people around me, at least.
People who are not so comfortable with english will _endure_ the brazilian portuguese if you provide no alternative.
And personally, I dont need people with masters degrees to translate for me. I'd much rather have people with good marketing skills/understanding and a solid language.
We have a web app that two translators use initially independently to translate our entire corpus. The app then asks them to select the best option where there are mismatches (they can also enter a third alternative). There are a couple of rounds of this.
Finally an editor has access to all rounds of the translation in an editable table (http://i.imgur.com/zPJp0sz.png) where they can see who has inputted what, and wether things have been marked as 'acceptable' or 'completely wrong'.
This works well on most of our short/simple strings.
It also helps that we have two voice actors who are also native speakers come in and record the script, where any further errors or better translations can come up.
We do this because we're translating with different goals than the majority of translation purposes. We need to focus on translating for speech at relatively low volume and for those learning a non-native language. Most translation shops focus on written language for native speakers, high volume, speed and cost.
What I do think can be automated though, is the act of finding search phrase equivalents for different languages.
And if you have some humour or poetic fomulations in your description, it probably wont give you a very good result.
If you want a hint of what Google Translate does to a language, do it full circle. Translate a text from English to for example French, and then input the result for translation back to English. The resulting experience is pretty close to what the French user would get.
(It also doubles as comedy.)
Then you've got all sorts of idiomatic expressions. It's often not as simple as plugging into Google. I've been working on iOS language apps for a few years. They are localized in a dozen languages. It's a lot of work. For example:
https://itunes.apple.com/us/app/h4-spanish-lite/id388918463?...
Is there any interest in maintaining an open source list of common words and expressions? GitHub? Google Spreadsheet? I could seed it with some of my data. I already have some common words for a dozen languages here:
https://docs.google.com/spreadsheet/ccc?key=0ArVkFagUZg7bdHB...
If anyone would like to contribute, let me know.
I can only speak for myself, but there is a special place in hell for any and every one who butchers my native language into an indecipherable mess and have the guts to serve it back to me.
There is also lots of other small details on how people are going to use your app. For example in some countries, the last name is supposed to be put before the first name and if you put it the other way around, you would still have ~15/20% of people which would put their last name on the first field anyway regardless how it's labeled.
Will update with the data once done.
But as David Janner says in the article linked to below, he did't get a lot of result from it. But if anybody knows any good translators to south korean, I'd be happy to include it in the article.