We have covered a number of themes on this blog to do with optimising website conversion rates, but we haven't talked much yet about geo-targeting and how it can improve your A/B testing and sometimes help you avoid serious mistakes. Here's a fun example to make you think:
Imagine that your site gets 2,000 visitors every day from London and 500 every day from Paris. You decide to carry out some A/B testing, so you show visitors two versions of your landing page and measure the conversion rates. At the end of the day, you find that version A had a conversion rate of 80%, while B had a conversion rate of 20%. Pretty obvious that you should drop B and use A, wouldn't you say?
Not so fast! In fact (because this is an illustrative example) that data was generated from a model where every single visitor from London converted if they saw A, while every single visitor from Paris converted if they saw B! In this example, both A and B are really excellent landing pages, just suited for different markets.
Any kind of segmentation of your visitor base can help you make more accurate website conversion rate optimisation decisions. The trouble is that, until you have some form of conversion, you usually don't have any information about what segment the customer belongs to! One of the few exceptions to this unfortunate truth is that nearly every visitor to your site gives away a little bit of information in their IP address. It's possible to use this information to work out what country or geographical location they are visiting your site from.
Sadly, the way that IP addresses are allocated geographically isn't straightforward at all. Geo-targeting is as much of an art as a science and relies on large databases linking IP ranges to physical locations. It works, though, and for a lot of optimisation situations, every little bit of data you can collect can help. It's not just a matter of customising local versions of your site for different regions – it can help with your conversion rate optimisation too.