A/B ( c ), As Easy As 1, 2, 3… A Guide To A/B Testing
He thinks the homepage’s Call-To-Action (CTA) should be green, but your certain it should remain orange — it’s a battle of wills within the office as both of you think you’re right.
But how can you actually determine what colour your CTA should be? This is an important decision to make, as a small change to a website can have a massive impact on its conversion rate — For example, changing the colour of the homepage’s CTA increased Hubspot’s conversions by 21% — so what is it going to be, green or orange? Fortunately, for the sake of peace within the office, A/B testing is your saviour!
So what actually is A/B testing? Think of A as representing your original-unchanged website (complete with its orange CTA) whilst B represents the newer version of our website (complete with green CTA). As the name suggests, the test compares the performance (in terms of conversion rate) of page A with page B, in order to determine which performs better — specifically, which makes your visitors more likely to become your customers. Half of your website’s visitors will be shown page A, whilst the other half will be shown page B.
After some time, you will be able to determine whether the change has a positive effect your visitors’ behaviour or not (and whether you should keep a low profile for a while as he gloats — sadly for you, the A/B revealed that the green CTA outperformed the orange, whoops… no one is perfect anyways…). By assessing one element of your webpage at a time, it enables you and your team to pinpoint which of the changes led to a significant improvement and which should be ignored (cough… cough an orange CTA…). You can then accumulate the successful changes in order to produce the best possible experience for your visitors — everyone loves a happy customer!
Don’t worry about you’re A/B test affecting your website’s ranking (phew), Google actually encourages the use of A/B testing! However, it is essential that you don’t use A/B testing dishonestly, specifically cloaking (displaying different content to visitors and the search engine) is prohibited by Google. But I’m sure you weren’t thinking about doing that — you just wanted to prove your colleague wrong.
This is all great… But how do you actually know what needs testing? These A/B testing platforms do a great job at conducting tests and revealing its outcome– but how do you actually determine what is worth testing? You’ve tested one CTA, but now what? Which component do you change next and on which specific page of your website? — Even once you’ve made these two decision, what changes do you actually make? There isn’t much point in running a test on something completely arbitrary — it’s not going to make a ground breaking difference to your website’s conversion…
So what now? You’ve got this amazing A/B software, but you’re up the creek without a paddle — you need assistance, support, anything… Fortunately, you aren’t alone anymore — Scenario is here. As well as offering A/B testing, this software places a huge emphasis on guiding individuals through a structured optimisation process. Specifically, it will identify and suggest areas for change on your website based on its audience, context and a library of 158 neuroscience principles — so you know exactly what to test and why you should test it!!
For example, why does the colour of the CTA on your homepage even matter (except of course in terms of office pride) — well, the ‘Von Restorff effect’ (1993) describes the tendency for individuals to remember elements that stand out; as the contrast evoked between one element and the others causes our brains to ‘wake up’ and pay more attention, — meaning that the different element will not only be noticed in the present moment but also stick in our brains for longer. At the moment most of elements on your website are already orange, which is why the green CTA is (sadly for you, but not for conversions) more effective — as the green creates a notable contrast against the orange.
It gets even better; you can implement these personalised recommendations without even touching the coding of your website (score!!) — so there is no need to bother Dave from the IT department (Hi Dave).
So, settle that office argument once and for all with Scenario by your side.