Why You Should Care About Web Translation
David Sowerby explains why web translations are so important to companies of all sizes in a connected world.
A couple of caveats.
1. This is going to look rather long on this page (probably 4 or 5 screens) so I apologise now for that. The content is available as a pdf (above) if you can’t be bothered scrolling down screens.
2. I wanted to call the title “if you are a company operating in a western economy (read debt ridden Europe and the US) then tackling web localization is a matter of when not why” but that was far too long for a short distinct page title that google will like. I am open to suggestions.
In the course of spending the better part of the last 2 years grappling with web localization and solving some of the emerging issues (and attending a large number of conferences in the process) we amassed a fair amount of information and experience on whats actually going on out there at both a macro (changes in trade flows – US based subsidies to generate export income etc etc) and at a micro level (lots of small to mid tier firms now realising that the only growth platform in town is actually out of town). So I promised a few too many people I would write it all down – here it is.
The Localization Industry – It’s Changing fast
The global financial crisis has accelerated the shift in the balance of power from the West to the East. It has also had the effect of forcing companies of every size and across almost every industry in most western economies to focus far more heavily on the emerging economies in South America and ASIA as the primary growth opportunity going into the next decade.
Growth Forecast for Localisation Industry
For the localization industry this should be great news. WIth more companies operating in the global marketplace there will be more content for translation, which means greater demand for their servies – but is that actually true.
With content is increasingly shifted into a digital context, consumed from a variety of channels, and the demographics of the internet audience globalizing at a much faster rate than previous estimates, the growth path for the localization industry is actually far less clear than it at first seems. What is clear though is that the way content is translated will be vastly different from the current model.
Factors Influencing the Localization of Content
The New Economic Realities
The BRIC countries represent the growth engine of the global economy. The opportunity these regions represent from a purely economic standpoint is enormous. As the emerging middle classes in these regions grow their spending powers and levels of disposable income will increase – leading to an increasing demand for western companies goods and services.
Breaking into these markets, however, represent an vastly different proposition than most western companies have faced before, and it is primarily down to the fact that despite the commentary English is not the lingua franca of the Internet.
Language and the importance of cost effective communication becomes a key pillar to successfully growing in these markets.
BEING FOUND (OR MULTILINGUAL SEO)
Being found…. If you have cracked the communication challenge and you are localising your content then it becomes important that your potential customers can fid you as you try and break open a newmarket. Pushing to one side the arguments around social tools like facebook, twitter et al. the majority of the global internet community use a search engine as their primary entry point to what they want to find on the internet. So it stands to reason that ensuring you can be found in the relevant regional search platforms is reasonably important.
SEO and MSEO (Multilingual Search Optimisation) is an incredibly detailed area in it’s own right, however, across regions Google doesn’t dominate ( in fact in key Asian countries like China, Japan and Korea, Google struggles for any meaningful marketshare) and understanding how to ensure you can be found across different search platforms is a continued driver to how much organisations spend on translation. The concept of content volume is king doesn’t hold true.
The importance of proper localization cannot be underestimated when looking at search rankings as a fundamentally important part of gaining traction in new markets
The Changing Internet
By 2015 Chinese will be the number one language spoken by online users. there are already 748M internet users in Asia (well beyond Europe and US) but penetration is only 19%. ASIA is only half the story though as the improvements at an economic and access level across South America, SE Asia and the East is seeing rapid growth in other languages – Spanish and Portuguese particularly. The fragmentation of the language landscape on the internet is going to continue to be a major driver in how the web evolves.
What this means for the localization industry is fundamentally that there will be an increasing demand for localization of content, that the average amount an organisation is able to spend to get content localized will not increase at anywhere near the rate that the volume of content to be translated will and that the number of languages content will be consumed in will increase substantially from the current state. Lets look at these points in more detail
Demand for Localization Services.
Common Sense Advisory puts the forecast CAGR for the industry at around 14% for the next 5 years or so. That number is probably not far off the mark, however the underlying components that make up the sum total of every dollar spent on localization is changing. Primarily the continued investment by technology companies (like google) in machine translation toolsets will see a far greater volume of words handled within the same (or similar spend). If you consider the growth rates of online content then the only conclusion that can be drawn from the current industry forecasts is that a). A lot of the content created does not have enough inherent value to have a human translator work with it and b). Machine translation tools will continue to play an increasingly important role in reducing the volume of words a human translator needs to work on.
The other factor driving demand is the increasing volume of mid-tier companies operating at a global level. Large corporates have run “localization” programs for a number of years now but it is the mid tier companies in the US and Europe (particularly) who now find themselves competing on a global playing field. If there is one near undeniable outcome from the GFC it will be that companies across all sizes and industries in the US and Europe will need to start building and exporting goods and services into the growth hubs of Asia and South America. You simply cannot expect to do business in these regions without engaging with customers, prospects and partners in these regions on their terms (which means – in their language)
Localization services have primarily two cost elements. The project management cost associated with managing translation projects and the cost per word to translate content. This pricing models is extremely effective. It makes the cost of translation jobs simple to manage, it is a simple, effective and easy way to track the cost of translating a piece of content from Language A to Language B. It is however also expensive. In reality a word (or group of words) has an inherent value based on the context they are being used in and the importance of understanding between the organization and the end user. Certain content has a much higher inherent value to an organisation that others. For very important content (legal, medical, product driven) paying a higher per word price is justified – for a lot of other content there is no economic basis for that cost structure. The reality on the internet is that for most content on the web there is a cost vs accuracy dynamic that will continue to shift to a lower price tier.
The challenge is therefore firstly for technology to enable a framework were the appropriate translation workflow can be matched to the organisations view of the value of translating a piece of content and secondly to have a pricing model that more easily reflects the increasing volume of content wanting translation (without imposing the stiff penalties per word pricing entails).
Our view would be that a per hour pricing framework is a much more robust pricing model for web content – particularly when the human is being primarily in a post edit framework.
The Relationship Dynamic
““If I’m selling to you I’ll speak your language. If i’m buying die mussen Sie Deutsch Sprechen”
This is a simple (yet fundamental) driver to the vast changes we are going to see on the web in the next few years. Historically Western economies were firmly entrenched on the Demand side of the Supply/Demand economic equation. Lower cost economies in Asia and South America drove the supply side.
For a lot of US companies the concept of Globalization meant breaking into Europe and running their web presence in primarily English (40%+ of Western Europeans can speak english as a second language) with some basic local language support, and as most US companies were buying from Asia and South America the relationship dynamic was firmly with them. Not anymore.
The burgeoning middle classes in China, India and Brazil and associated rapidly increasing consumer and business spending power of this group means the buyers of goods and services in the near future (if not now) are firmly not in Western economies. Quite simply if as an organization you want to sell into the growth markets you need a localization strategy – expecting to compete in local markets with your core information platform illegible to your target audience does not make sound strategic sense.
Mobiles, Desktops, pdf’s, tablets, netbooks, laptops are all information consumption points. 5 years ago you publish something on the web and your audience could consume it. Not anymore. 350M smartphone users, offline/online desktop applications, Rich Media, Web Browsers, Offline Content – the sheer number of areas that a user can access content from places more demands on localization services. Were content lives, how easy it is to access and how quickly that content can be shifted to different channels is now a major advantage/disadvantage for any company operating in a multilingual environment. The logical conclusion is that (at least for public facing content) an organisations information should be authored, maintained, translated and published in the most accessible medium – the Internet (or in the interests of using more jargon – The Cloud)
The Pathway Forward
Languages (and translations) will become progressively more important for the vast majority of organizations – that I think is undisputed. Translations will become more democratized as the evolving models of crowdsourcing. machine tools and the commoditization of translation continues at pace.
So for organizations considering their first run at web translations or looking to reduce the budget impact of their existing multilingual projects there are some key trends and likely advances in technology that should be taken into consideration
Technology will ultimately eliminate the “middle men” who’s primary function is to shift content from one system to a translator and back again. This process is both inefficient and costly adding little value to the overall translation process. Eventually the translator needs to come to the content (not the reverse) for an efficient translation supply chain.
- Computer assisted translation tools will be the way forward for the localization industry. Human translators primary role will more and more be as a post editor of pre-translated content – the economics dictate this has top be the case.
- Human translators are vital in ensuring a quality output. Machines (for the foreseeable future – note I will not be taking bets that the team at Google will not prove me wrong) will never get to 100% accuracy. Humans are going to be needed for some time if the translation output needs to be more than basically legible.
- Search remains a vital marketing tool. Google is not the only game in town and your search strategy (and tools you use) need to be focused on the different requirements of different regions
- A per word pricing model makes no economic sense for the person paying the translation bill. If you believe machine tools can get progressively more accurate then a per hour pricing model can deliver cost efficiencies that reflect these improvements.
If you want more information on anything above or interested in a discussion about the points raised here contact me at email@example.com (or on Skype david_sowerby)
Would you like some proof we are significantly faster and better value?
The charts below show the quoted time (days) and price to deliver a 25 page technical manual from English into the 12 most popular languages (French, German, Chinese, Japanese, Italian, Portuguese, Spanish, Korean, Arabic, Dutch, Russian, Indonesian)
In terms of cost for the same quoted project, our average cost per language was 40% lower than the nearest competitor and 300% less than the most expensive quote.
Straker Translation were a massive 8 times faster than the slowest alternative quote.