In today’s post I thought we’d dive into the teaming waters of website testing, with the wonderful Justin Rondeau – Chief Evangelist & Editor of WhichTestWon, the top weekly publication in conversion optimisation.
So one of the things I find particularly interesting (and at times very frustrating) with regards to testing, is the fact that so many businesses simply aren’t making the most of this really useful, valuable process.
If I told you that you could double, or even triple your sales through your website, you’d jump at the chance, right?
Well, given that testing is one of the best (and actually really one of the only) ways in which to do this and learn from your results in the process, it seems crazy that we’d look such a winning horse in the mouth.
When I asked Justin why he thinks testing isn’t adopted across more businesses, he cited a report they released just a few months ago in which they discovered two top reasons:
It’s sad to say, but often times when the will is there, the people championing the cause simply don’t get the buy-in from management.
Which means that one of their best hopes for improving their site and increasing conversions (and ultimately, making more money – which is what most businesses would like), gets lost because of the culture of the organization.
It’s not a particularly easy problem to fix, however showing your managers how much they’re currently losing by ignoring this strategy can be a psychologically compelling way of re-framing the issue, so that they consider your proposal and (hopefully) change their minds.
In psychological terms this is known as ‘loss aversion’. The fear of losing something we already have (in this case, revenue), is more powerful than the thought that we might gain something in the future.
So, rather than making the case that by testing your site you could gain x amount of revenue per month, instead show them how much they’re currently losing by not allowing you to run your tests.
That should help shimmy things along a bit.
The second reason that you may not be conducting tests on your website, is a lack of resources. Maybe you don’t have the developers you need to set up the test and create the content, or perhaps you don’t have access to copywriters to help you change your text?
Well, as it turns out, the barrier to testing has really plummeted in the last few years, and while you may not have access to the team you need, there’s no reason why you can’t start testing some of the basics now, using inexpensive, but effective tools.
I personally love Optimizely - as a starter’s toolkit it’s fun, easy to use, and requires no coding skills whatsoever. Even if you just spend a few hours a month designing tests and experimenting with a variable or two, just by getting into this mindset and getting familiar with the process will help you gather useful data so you can hit the ground running when eventually your resources are in place.
If you’re new to testing, getting started can seem like a scary, unfamiliar and daunting prospect. Justin gives some great advice as to where to begin:
"I generally go with the areas of most opportunity, and to me those are the areas that are actually at the bottom of your funnel"
Why? Because these are the people who've made it the furthest into your site - they've shown interest, jumped through your hoops and interacted with your site to the point that they're already invested.
“You don’t want to lose these people – these are the areas of highest opportunity, as well as areas that are going to be a lot easier to measure the ROI.”
If you’re swimming around in the deep end (in this case, deep referring to the fact that these peeps are invested), you’re not going to have to worry about trying to work out how much a click is worth in monetary terms for those people at the very top of your funnel.
Instead, you can focus on designing tests the results from which you can use to generate the biggest impact, helping you to optimise your process where you stand to gain the most (scary as that may sound).
From a psychological perspective, because these customers have already been through (what Justin describes beautifully as) the ‘site gymnastics’, chances are they’ll be a bit more forgiving too if the variables you’re testing don’t work out with an uplift the first time around.