If you want to know when new articles go
subscribe to the WebWord.com
to Localise Your Website:
A UK Perspective on Usability
Travis (WebWord Guest Author)
It's often said that a common language divides America and England. If an American says "I'm mad about my flat", there is something wrong with his car. If an Englishman says it, he likes his apartment. If the Americans and the English can't communicate without ambiguity, what happens when non-English speakers use your site?
One frequently-cited example is the fact that the word "Sale", although eye-catching and attractive to an English speaker, translates as "Dirty" in French. Hardly the kind of advert you want to splash over your home page.
System Concepts is a 19-person usability consultancy based in London's Covent Garden. Over the last two years we have seen an unprecedented demand for our services in usability. We think this is for two reasons. First, there just aren't enough usability professionals to go around. For every qualified usability professional, there are currently around five jobs on offer. (One result of this is that many of these positions in the UK are being filled by unqualified people re-inventing themselves as "customer experience architects", but that rant is the subject of another article). The sheer lack of qualified people means that people turn to consultants such as ourselves.
The second reason is that many US companies don't have the resources to usability test in Europe. If it was just a case of turning up with a computer and a video camera then anybody could do it. In reality, you need to find the most appropriate city for the target audience, recruit participants and pick the right incentive, set up the venue and get your Internet connection to work (even with the right adapters this is never trivial!), then run the testing in the local language. Facilitators need to be properly trained: some cultures are reluctant to criticise, and a good facilitator can draw out issues that might otherwise be left unsaid. Our usability experts include native speakers of Spanish, French, Dutch, German and English which means we can cover virtually all of Europe, and certainly the top Internet-using countries.
If you're toying with the idea of making your site accessible to non-US cultures, here are some tips from our experience. You can
To succeed in Europe you will have to localise
Pro Active's "Pan European Internet Monitor," a survey of more than 150,000 Internet users in 15 European countries shows that 82% of European online users can read English. But
65% of those surveyed prefer sites in their own
The bias even extends to the level of the domain name. The "dotcom" moniker conveys uniquely American content. Evidence for this comes from the fact that localized Yahoo! Web sites are amongst the top five sites in France, Italy, Spain, Germany, Norway and the U.K. Other top sites in these countries are usually local sites or country-specific versions of Yahoo!'s portal
Don't translate, localise!
Most of our clients appreciate that localisation is different from simply translating your website, that is making the words suitable for the local language. With localisation you need to consider broader aspects of your communication. You need to get the visual elements right, such as the right icons and the right formats for things like dates. But you also need to re-think the underlying content and usage scenarios, which may be different. For example, some cultures do not like the idea of home delivery (they prefer to collect purchases from a store). Translation can be done by a bureau in the US, but to localise properly you will need to do some research and iterative testing "on the ground" with the appropriate audience.
Test a wireframe site
Many of the sites that we have tested are very large, with thousands of pages and considerable dynamic content. How do you go about internationalising such a site, let alone test it? Our experience shows that it doesn't make practical sense to test every page. Instead, test a wireframe of the site: this is a top-level view of the navigation framework. In our tests, a typical wireframe site comprises a fully-translated version of the home page and uses all the standard visual elements from the original site. It also has fully-translated versions of the sub-pages needed to kick-start typical tasks. But the detailed page content (for example, product information) is greeked. This saves an incredible amount of time and means that testing can be carried out very early because only a small subset of the site needs translation (80% of the translation effort is spent on the detailed sub-pages). It also means you identify the most important usability bloopers: for example, using the wrong term for a navigational item on the home page can disrupt many tasks, but a similar error on a product page will usually affect just one task.
Compute the cost-benefits of localisation
One of the advantages of testing a wireframe site is that you can work out the cost benefits of building a localised site before you have invested too much money in the process. With a properly-executed usability test you have real, concrete customer data on which to base your judgements. These measures might include the conversion rate, task completion times or the ratio of successful purchases compared with a localised competitor site.
Example: Internationalising a Global dotcom
We recently completed a major usability evaluation of an e-commerce site hosted by one of the world's top three computer companies. As well as selling computer equipment and peripherals, the site also provides an extensive database of customer support material (such as manuals, FAQs and driver software) that can be downloaded from the website. This case study serves as a useful example for some of the points made above.
We were commissioned to explore the return on investment and customer loyalty benefits of localising the web site in France and Germany. The work entailed recruiting representative participants in Paris and Hamburg, equipping usability labs in both cities, developing realistic customer experience scenarios, carrying out the tests in the local language, and designing and analysing the data from the testing. Deadlines were such that initial feedback was required within 24 hrs of test completion.
To control for practice effects, we used a between-subjects design: half the participants carried out the tasks only with the localised site and the other half carried out the tasks only with the English site. Language ability and web experience were matched on a per-participant basis in the two groups.
The usability testing clearly showed significant benefits from localisation, in terms of customer effectiveness, efficiency and satisfaction. However, the decision to localise was a business one, based (among other things) on the likely revenue increase to the client. We therefore used the data from the usability testing to provide the client with the information needed to make this decision.
Assume that 1000 people per week, based in Germany, visit the website to buy a computer peripheral valued at $200. Our usability testing showed that, with the localised site, 89% would be successful. In contrast, if the localised site did not exist and customers had to use the English site, 64% would be successful. This is a difference of 25%, or in our example, 250 customers. Assume that these customers choose instead to go to a local electronics store to buy their printer. Not all of these customers will still choose the same brand. The store may have a special offer on a competitor's product or the customer may be sufficiently disappointed with his experience of the client's website that he decides to switch brands. Let us make the conservative assumption that the client loses 10% of sales due to this effect.
These assumptions lead us to estimate that the client would lose 25 customers per week, or in our example a revenue of 25 x $200 = $5,000/week. This scales to $260,000/year. Clearly this value will be magnified significantly if we consider the full range of the client's products.
Based on our report, the client decided to localise both the French and German versions of its website.
About System Concepts
is one of the UK's leading consultancies specialising in usability. We were founded in 1981 and have provided a diverse range of services to an extensive client list of blue-chip organisations which value quality. We have carried out usability evaluations of various e-commerce sites throughout Europe.
All our consultants have postgraduate or Ph.D qualifications in the behavioural sciences, enabling us to provide scientifically valid and credible results. With their top-notch people skills, our test facilitators engage participants fully in the evaluation making sure that all their concerns are uncovered.
Through partnering arrangements, we also have access to testing facilities across Europe and have carried out tests in Germany, Spain, France and Holland.