Return on translation - a partnership equation

The case for it is pretty solid on the whole.

Research from Common Sense Advisory shows that a typical investment in translation and localisation is minuscule compared to the revenue it enables. The budget for these services generally represents less than 1% of marketing or R&D budgets, even when staffing and technology costs are included.

In a survey of Fortune 500 companies, CSA further found that companies that increased their translation budgets were 1.5 times more likely to experience an increase in total revenue.

Understand your market

But ROI is influenced by a wider range of factors than the cost of translation alone, including the cost of regulatory compliance and countless other costs of doing business in a new country. And it's different for everyone.

Microsoft makes a large chunk of its revenue in non-US territories, and translates into more than 90 foreign languages and language variants. Toshiba, too, considers it vital to translate (30 languages), as does Apple (around 40 languages).

Unexpected pay-off

Ultimately, the most useful way of approaching the problem may be to consider the opportunity cost of not translating. And that depends entirely on the product and how it taps into each new market.

Even the most experienced campaigners don't always anticipate the nature of the reception they're going to get in a new market, or indeed the type of ROI.

When a top-tier software vendor entered Brazil, the Portuguese translation of its Web design suite sold few copies, and was widely pirated. Does this mean they blew tens of thousands of dollars? Perhaps, but in reality the lost revenue is best characterised as 'school fees'. The company hung in there and used the exercise to polish its copyright protection, with knock-on benefits in other markets. Presumably, higher sales resulted

Another client's experience with Brazil and China was similar - it combated a lot of piracy but the market size made the foray worthwhile. Not translating would have left it worse off.

Best approach

When in doubt whether to translate, test the waters with a measure of localisation - perhaps starting with product literature and support activities - leaving the option of localising the product later. If a spike in sales results, translation may be a component of the sudden success.

Also, understand your clients, including the languages they speak and media they consume. In South Africa, English is the main business language, but by sheer force of numbers isiXhosa and isiZulu are just as important (in a commercial sense) to cater for.

Look for similarities in other markets. Uganda, where Luganda is more important than English, is one example, and translation has served clients well in that territory. More fragmented markets (like Ethiopia, with 75 languages) have a more difficult (but in principle similar) decision to make.

Choices, choices

The next question, of course, is what type of translation to invest in. This one may be a little harder, and it further complicates ROI considerations. Should you run it through Google Translate - a 'best-effort' service that is nominally free (up to to 100,000 characters per day, at any rate)?

Are you inclined to trust human accuracy but tend towards a cut-rate deal with minimal vendor management? Or do you partner with a language service provider (LSP) that offers the best of human accuracy and machine efficiency, coupled with business process engineering and project management to help you attain maximum translation efficiencies in the long run?

This depends to no small degree on how you view translation, but ultimately also on your own circumstances. If multiple language translations will become necessary over time, high volumes are involved and regular updates are par for the course, doing it right from the start is key. Investing in new processes to govern content generation and translation as well as your relationship with your LSP partner is your best chance of attaining consistent quality and achieving progressively more cost-effective translations over time.