How to optimise

In this global economy, online businesses are expanding internationally faster than ever before. Internationalisation, for all the additional complexities it introduces, is a blessing for companies engaged in optimisation programmes.

First, multiple sites in multiple languages increases the flexibility and amount of testing one can perform. Let’s consider a retailer with a single language. Testing will focus largely around the central goal of sales. 

This can handcuff you in terms of the number of tests you can have running at any given time, as there will be a conflict between tests (skewed data). Multiple language sites afford you the option of alternating tests between your primary markets or languages. No more conflict.

Second, you can create an opportunity where none exists. If you have presence in multiple countries, you may struggle with volumes in certain markets which are then deemed unsuitable for running tests.

To solve this problem, consider running a test across multiple markets of the same language. For example, Germany, Austria, and Switzerland all could have their own German-language sites. The cumulative traffic across these three sites may be just what you need to allow you to achieve a statistically significant result. 

This may contradict the idea that different cultures should be treated differently, but think about it this way: before, you had no testing opportunity from any of these three markets. Now you have one.  Just like pancakes and (Canadian) maple syrup. One is nothing without the other.

One further strategy would be to simultaneously test common elements across multiple sites. If you run the same test at the same time segmented by language, you can achieve multiple winners.  Make sure you stick to testing non-content related elements such as styles or layout to ensure planning and development is minimised, while results are maximised.

What to optimise

Approach your testing of international sites in much the same way you approach your testing today: look at what your existing data and studies tell you and formulate a hypothesis.

That doesn’t mean we haven’t been exposed to a few interesting trends. On French retail product pages, we have observed that the equivalent of ‘Buy’ or ‘Buy now’ can outperform the more typically English ‘Add to Cart’ or ‘…Basket’.

‘Buy’ is what we would deem a ‘hard’ call to action and often loses you visitors early in a funnel because of the sense of commitment it provokes. At the same time, it’s more likely going to weed out the less qualified visitors and can drive more qualified visitors down the funnel. This is the effect observed on two French sites.

If you are testing on a Japanese language site, your approach will undoubtedly be different. A few years ago we saw that a more playful, but very wordy version of a customer’s landing page was responsible for significant performance increases.

Both of these examples are themed around messaging. For internationalised businesses, we often see a difference at this level – whether how you emphasise free delivery or how you persuade visitors to apply for something. 

Lastly, what are your geo-targeting opportunities? Are your visitors even arriving at the right site?  You’d be surprised at the number of sites that struggle with this. This is something you could easily solve with your optimisation solution.

Furthermore, regional segmentation by geo-IP is becoming more and more accurate, so there can be plenty of optimisation opportunities when it comes to offers based on region.

That’s it for now. Check out the concluding part in this series where I will discuss the challenges of brand consistency when optimising globally as well as my final thoughts on the topic. It’s a complicated process but a worthwhile one, I promise.