Conversion rate optimisation is about targeting as well as testing

Every time I read one of the results of the survey’s that eConsultancy does, I am more amazed than the previous time I read one. This week we’re looking at the Conversion rate optimisation report.

As companies get better at conversion rate optimization (CRO), it is becoming harder to improve conversion rates, with 65% of companies see improvements in conversions, compared to 70% in 2009 and 2010.

Although more companies are doing conversion rate optimisation, a smaller percentage than ever are seeing improvements. Is this because they are doing it wrong, or is it because they’ve reached the edges of what is possible? My suggestion is that they are reaching the edge of what is possible.

Although many people are missing a trick, because one person does conversion rate optimisation, another person is doing campaign management optimisation and another is doing landing page optimisation. All of these people are adopting the methods that eConsultancy mentions (A/B testing, copy optimisation, etc), but they are not looking at how they do them all in conjunction with each other.

As an individual I do lots of strange things on websites and it is almost impossible to tell from this what was going on in my head. I sometimes use tabs meaning that my journey may make little sense if you monitored it from start to finish. However, the beauty of web analytics systems is that it can aggregate a load of information about different people to get a useful impression of what is going on.

For example when doing my form optimisation I’ll look at common drop out points in the process. For my multivariate testing I’ll try different things on those common drop out points to work out which one works best to improve the total conversion for everyone.

For my landing pages I’ll try different set ups to see which one provided the biggest click through rate and conversion. I’ll look at A/B testing different pages to get the best results.

For my campaign tracking I’ll be looking at open rates, click through rates and costs per click. I’ll be working out how to get the most people to the site for the best revenue, by looking at ROI for conversions.

Really though I should be doing all three in conjunction – I shouldn’t have one landing page for lots of campaigns and I shouldn’t have one form for all my users. This means not just a lot of testing, it means a lot of targeting. One of the real advantages of a tool like Adobe’s test and target is that you have all your web analytics data straight to hand and can do the actual targeting in a very easy way.

This is the real test of your organisation’s set up, however. Many organisations have a number of individuals who do web analytics reporting and a number of individuals who act upon the data as part of their job. Having someone whose sole job it is to come up with new ways of targeting (and testing within that segment) based on web analytics data and then advising the other teams (marketing, campaigns, business, design, etc) on what to do is a new way of working for many organisations and needs lots of buy in from up high otherwise you won’t get the best return on your investment of the tools.