Date: 12th June 2015
In Part 1 of our blog on taking your e-commerce global we covered infrastructure expansion. Here in Part 2 we talk about considerations for localising your storefront and how best to mitigate the risks before going live.
International usually means not English – Does your site read well in the target local language. It is obvious, but the subtleties of language are many and web copy has just a few seconds to engage or lose the reader. Use a local team to help translate your pages. Check for any degradation of your page appearance speed as some languages have longer sentence structures than the English version, it can be anything up to 50%.
Be prepared to throw away your hard-won Usability UX! – To be successful you need to make your site attractive to your new users. This may include adapting your pages to fall in line with local web-store practise: you may need to abandon some of the look and feel of your original site – maybe weird colour combinations, changed CTA button positioning – for some of your team it may feel like going back 5 years in UX practise!
It is all about the money – If you are going to use currency conversions for new markets, how often will your prices track currency fluctuations? Can your platform handle different, regional promotions, delivery-charge algorithms per country and within a country?
If you have a lot of personalisation on your pages supplied by third parties, you will need to make sure their services work in new territory and the information provided is relevant to the locations audience.
Be a Boy Scout
Benjamin Franklin once said; “By failing to prepare, you are preparing to fail” and it is critical for any eCommerce expansion programme to make sure your new offering is thoroughly tested. The testing you undertake should focus on creating the ‘real’ customer journey flows.
Take a Test Drive before the big day. Load test your key journeys across a full range of devices (static and mobile) well before going live. This will ensure your online store will meet user demands and all key functionality works as expected against realistic user journeys. Do not let stretched tech-teams decide what a ‘realistic’ load mix is – under time-pressure you will end up with an artificial test whose numbers are less helpful than not knowing! You should run tests until your system ‘fails’, it’s better to find it out in staging if you have enough capacity to match the most optimistic prediction beforehand rather than watch it fail on the big day.
Your testing will also need to test boundary conditions/unexpected operational anomalies such as:
- Promotions are only displayed in correct locations – UK BOGOF not displayed on another country’s pages
- Reported stock levels are correct for the local logistic commitments (next day delivery, etc.)
- Prices are correctly displayed based on location & currency
- Correct Images are being pulled from the appropriate Cache or CDN to avoids any cultural sensitivities
- Delivery charge calculation algorithms are working correctly based on location and distance from the warehouse location
- Third party sources operate in your chosen new territory; failure to display correct information based on location information
There are more Big Days than the first big day. Yes, you have successful handled the ‘big’ launch – effective campaigns have been driving high volumes of traffic, and your technology has been delivering a fast and smooth user experience to hordes of new visitors.
However, you need to keep generating sales week in week out, which is when you need to be changing things fast: fixing bugs, moving layouts, A/B testing, refining features. There is a huge risk that those changes damage UX with slowdowns and new errors. There will also be events like Cyber Monday, Christmas and other seasonal activities that can unexpectedly see high demand.
How do you manage the situation? You need to monitor 24/7 key journeys pre- and post- launch to identify and resolve performance issues before they affect your customers. You want reports focused on the issues affecting the customer journey rather than pages of potentially false positives from conflicting infrastructure alerts.
There is huge growth potential in globalising but it can go wrong if attention is not given to maintaining that initial customer journey performance.
CDN: Content Distribution Network: You can improve global availability with CDNs. Customers far away from the servers will get slower journey speed (milliseconds here and there can quickly add up to seconds for a content-heavy web page), so a global content delivery network or cloud networks save those seconds to get their content closer to customers.