The tribulations of worldwide websites

Most large organizations need to serve local audiences online – but they also need to save money, says David Bowen. How can the two be reconciled?
 

I have been reading a PhD proposal headed The Phenomenology of Subjection. I don’t understand many of the words, let alone the sentences. But at least the headline warned me that my brain was about to be stretched. Less dangerous perhaps than something that sounds simple, but is far from it. This, for example: ‘What should large organizations do about country websites?’

It is a question bosses find easy to ask, perhaps followed by ‘Do we really need them?’ Corporate websites are bad enough on their own – they do not generate money, and it’s hard to measure what benefits there are. Country websites have the same problem, multiplied many times. Unless of course they are mainly there to serve customers, in which case there is a different question: Why bother putting anything corporate on them – it just gets in the way? There, an extra complication before I have even started.

But the boss is asking a good question – a useful starting point for almost anything is ‘Do we really need it?’, because it forces us to think hard. And there will be some who, having considered all the options, will be able to say, ‘No, we don’t need any country websites. Let’s go to the pub.’ But they will be a minority. For most the options are far more nuanced.

The ‘no needers’ are those that do not operate internationally – most US retailers, for example, as well as many Chinese groups. They may well need to talk to investors around the world, but they can do that in English on their corporate site.

At the other extreme are companies that sell to consumers around the world, so it is not a question of 'if' but 'how'. IBM has sites for Aruba and Burkina Faso; but they are selling sites, so they pay for themselves.

Most are somewhere in between. One group that might be thought likely to go the pub: business-to-business operations dealing with customers happy to work in English. But looking at examples, you will find they have all decided to offer localised material. Rio Tinto offers Japanese and Chinese sites, because English is not widely spoken there; Goldman Sachs used to do that but now even covers countries where English should not be a problem, like Germany; BAE Systems has an Arabic site to serve its biggest customer, Saudi Arabia.

It’s interesting that Goldman has moved from translating only where it is strictly needed to a broader approach. But even a German fund manager who works mostly in English will be a little more comfortable, and so receptive, if a bank does him the favour of speaking to him in his own language.

The pressure against this is of course resources. One multinational declared several years ago that it was going to cut its languages back to six. It still has many more than that, simply because it realized it would have risked losing business and goodwill by cutting out the ‘small’ languages. If you add the Hungarians, Czechs, Poles etc together, you will get a lot of people.

And of course it’s not just about language. To revive a cliché from the past – think global, act local – you are doing yourself no favours by coming across as an arrogant multinational.

Which is why most of the clever thinking now is about how to provide a local feel all round your markets at the lowest possible cost. Here are some of the options:

  • Country sites that share as much as possible. Look at Unilever.com. Now look at Unilever Argentina and Unilever Pakistan. They are all built on the same platform, and share words (translated where necessary) and pictures where they can. Unilever serves some smaller countries with a single site (such as central America and the Middle East), but there are still more than 60 different sites. This requires very strong governance, and relentless training, but it works well. It can be tricky to get the balance between localisation and efficiency right. For example Shell's otherwise impressive estate has a China careers page without a single Asian face on it, while Nigeria, where it is a big employer, has only one African on the same page.
  • Building country pages into the main site. Philip Morris International covers many countries in their own languages, using short web pages and downloadable documents (including a universal one on the dangers of smoking). Other languages are often available by clicking the selector at the top of the page: see Slovakia and Senegal for examples. Statoil also has extensive country information, though the great majority is in English only, which seems unwise.
  • Tailoring the approach to the country. Companies with country sites can group them into bands, and expect and support different levels of cover depending on how important they are. This may mean providing only contact details for some countries. BP is subtler, with an interesting hybrid approach. Its global page lists a large number of countries. Some, such as the UK and Trinidad and Tobago, lead to full sites. India looks like a full site, but has relatively few pages of its own. Other country links lead to a summary page with links to local websites, which might well belong to the subsidiary Castrol, to relevant career pages, or even – for Vietnam – to a local Facebook page. These are mostly in English but some, such as the Czech Republic, are bilingual. 

Which way works best for your organization will depend on many things – but these points strike me:

  • For Unilever, the key to success appears to be training. If you have people on the ground you can trust, and who understand your CMS, it all becomes so much easier.
  • BP's pragmatic approach will work best where you trust people to say what they cannot as well as what they can do. Why can Trinidad and Tobago run a full site when India cannot? I'd guess it's to do with the commitment of local senior management. So we can add 'education' (of bosses) to 'training' (of the people who will do the work).
  • You will also have to rely on local managers to work out what does and does not need to be translated – though your own budget may have the last word. In general, assume it will be difficult to cut languages; and it may well be a false economy anyway.
  • It's really all about governance. As ever.

- David Bowen