home contact sitemap polski

Translating Tumblr

April 23rd, 2012

In the past 6 months I have disappeared off the surface of this Earth and started to live virtually in the Blogosphere. All of this, because of my last large translation project – Polish localisation of Tumblr! I have to say that no other translation job has had such an impact on my life as this one and I can confirm with all certainty that Tumblr is very addictive!

What is it like translating a popular site like Tumblr?

Well, it’s been an exciting but also a humbling experience. Nothing puts you more under pressure than a few very vocal and devoted users of the platform who all have very clear ideas (often very different ones too!) of how things should be translated – if translated be they must!

I think what surprised me the most was the possessiveness (I can’t think of another way of describing this) of those Polish users, who felt that their best-kept secret was just about to be revealed to the wider masses of the unworthy. They have been using Tumblr happily in English and didn’t want to share the platform with a broader Polish audience. They liked the idea of being privy to something very special and exclusive (only in terms of language accessibility – Tumblr is anything but ‘exclusive’).

Having been on Tumblr for a few months, I think I can understand this now, but don’t appreciate the selfish viewpoint of those few. What I have learnt about Tumblr is that it is full of very creative and talented people from different walks of life and I get inspired every day looking at what they produce, discover and do. Some of them pride themselves for being an anti-mainstream crowd. They put themselves in opposition to Facebook, which is for “everyone else”. In this respect, there was some discussion in the translation process about how NOT to sound like Facebook!

I have also found some very generous users and have made friends. In particular, a wonderful person from Poznań, who was one of the beta testers, but turned out to be more than that. She kept my spirits high in the moments of “doubt and despair” and provided ample input and advice. Ave – I do hope we meet in person one day, as I still owe you a few drinks for keeping me sane back then!

Translating Tumblr is not a one-off thing. Tumblr is dynamically growing, changing, expanding and there is always plenty to do for the team of localisers (currently 8 languages and more coming up). So I have been sucked in, but sucked into something that I wouldn’t want to go away from…

You can find my Polish Tumblr blog here: Polka Dot. I also work on the Polish version of the Tumblr staff blog – Tumblr: Blog ekipy.

I have been recently publishing some articles in another blog on multivariate testing for international website optimisation.

I do not want to republish the articles, but I thought it would be good to provide just a little explanation and link to them here.

I have been interested in cultural usability since my postgraduate studies at Sussex University (in Human-Centred Computer Systems) and have done some research in the area of website design for different cultures. This has brought me to the technology called multivariate testing – which basically allows you to run several variations of the same website simultaneously in order to test which version provides best results (either in number of successful transactions, sales conversions or newsletter sign-ups, etc.).

Together with a few other researchers we have been investigating the application of multivariate testing specifically for website localisation – a brand new approach altogether.

So, for anyone interested in cultural usability or international design, I would invite you to have a look at the blog – www. Cultural Multivariate .com.


Image courtesy of Marcelo Bello

Image courtesy of Marcelo Bello

“Although there is evidence of convergence of economic systems, there is no evidence of convergence of people’s value systems. On the contrary, there is evidence that with converging incomes, people’s habits diverge.”
(de Mooij, 2000)

It is a common belief that globalisation is causing the world to become a uniform and homogenised place, and that people all over the globe will soon be behaving in a similar way. This view is not shared by researchers such as Marieke de Mooij, who insists that cultural values and traditions are deeply rooted and will not be easily changed or affected by globalisation (2000). De Mooij reasons that the increasing wealth in the world will bring a diversification of tastes as consumers will be able to better afford their personal predilections.

De Mooij’s prediction seems to be finding its confirmation in software and website designs, where the possibility of personalisation has become more and more prevalent (My Yahoo!, My EasyJet.com, Google Personalised Home, etc.). But surface customisation, such as offering several ‘skins’ or preferred settings and colour schemes for the interface are not enough when it comes to providing a localised version of the product. Localised means ‘as if made locally’ and a successfully localised product should pre-empt the user’s cultural expectations of what it looks like and how it functions.

It is not a surprise that many studies prove that people prefer to shop and interact with websites in their own language. One such cross-cultural study was conducted by Singh et al. and compared user’s attitude towards the site with their intention to buy (2004). The results of the survey were in favour of cultural customisation. A study conducted by Forrester Research (quoted in Singh et al. 2005) suggests that “non-English-speaking users stay twice as long on localized websites as they do on English-only web sites, and business users are three times more likely to make purchase online when addressed in their local language“. Such evidence should encourage service providers to ‘speak the customers’ language’ and adjust to their needs and preferences as dictated by the local culture.

Globalisation of services made possible thanks to the internet is a great opportunity, but also a challenge. As pointed out by Singh & Pereira “the Web allows companies of all sizes instant global reach and the immediate ability to interact with customers all over the world. (…) However, these advantages come with a price: a likely flood of competitors, all exploiting the same advantages, vying for the same target markets” (2005). The debate on whether to standardise websites for international use or adapt them to the target market is ongoing, but studies seem to suggest that in order to win customer loyalty and trust, companies need to compete at all levels: prices, quality of products, and cultural tailoring of their website and customer services (Simon 2001, Singh, Zhao & Hu 2004, de Mooij 2000).


De Mooij, M. (2000). The future is predictable for international marketers. Converging incomes lead to diverging consumer behaviour. London: International Marketing Review. Vol. 17. Retrieved June 20, 2006, from http://www.mariekedemooij.com/articles/demooij_2000_int_marketing_review.pdf

Simon, S. J. (2001). The impact of culture and gender on web sites: An empirical study. ACM SIGMIS Database. New York: ACM. 32, (1). p.18-37

Singh, N., & Pereira, A. (2005). The Culturally Customized Web Site. Customizing Web Sites for the Global Marketplace. Butterworth Heinemann

Singh, N., Zhao, H., & Hu, X. (2004). Analyzing the Cultural Content of Web Sites: A Cross-National Comparison of China, India, Japan and U.S., International Marketing Review, 69-88

Localisation of a product is often wrongly seen as equivalent to translation, which takes place towards the end of the development process. Regrettably, at this stage, it is too late to make any major changes to the underlying metaphor of the design, which may be completely alien and illegible to the target culture.

Website customisation for a target culture has to take into account many elements of the design at various levels. This would traditionally include: information architecture, theme and navigation, graphics, photographs, text translations, search engines, audio and video.

Alvin Yeo introduced a division of culture-dependent elements into overt and covert factors which need to be considered during the software/website localisation process. ‘The overt factors are tangible, straight forward and publicly observable elements. (…) Covert factors deal with the elements that are intangible and depend on culture or “special knowledge”‘ (1996).

• date, calendars, weekends
• day turnovers
• time
• telephone number and address formats
• character sets
• collating order sequence
• reading and writing direction
• punctuation
• translation
• units of measures and currency

• graphics/ visuals
• colours
• sound
• functionality
• metaphors and mental models

The accuracy of the covert elements has proven often decisive for product acceptance in the target culture, as presented by various studies for instance by Russo and Boor (1993), Zahedi, Van Pelt and Song (2001), Evers & Day (1997).

Unsurprisingly, it is the covert factors which cause a lot of problems for designers, as by definition they tend to be more elusive and difficult to identify even by people native to the target culture. It is however the covert factors that allow the design to reach the highest level of localisation: cultural customisation.


Evers, V. & Day, D. (1997) Role of culture in interface acceptance. Proceedings, 6th IFIP Conference on Human Computer Interaction. (July 14- 18) Sydney, Australia.

Russo, P., & Boor, S. (1993). How fluent is your interface? Designing for international users. Proceedings INTERCHI ’93 Conference on Human Factors in Computing Systems: INTERACT ’93 and CHI’93. Amsterdam, 24-29 April (pp. 342-347). ACM Press.

Yeo A. (1996). Cultural user interfaces: a silver lining in cultural diversity, ACM SIGCHI Bulletin, v.28 n.3, p.4-7

Yeo A. (2001). Global-Software Development Lifecycle: An Exploratory Study. CHI2001. Volume No. 3, Issue No. 1

Zahedi, F., Van Pelt, W. and Song, J. (2001). A conceptual framework for international web design. IEEE Transactions on Professional Communication. 44, (2), p.83-103.