To start, just take some time and review the image below. It’s a screen capture from a client’s recently completed split test experiment from their Google Website Optimiser account. If you look closely you will see Variation 1 of their home page achieved a 50% improvement in conversion compared with the original. Google’s tool is quite confident this experiment reflects the new page’s long-term ability, giving it a 96% chance to beat the original.
Hands up who else would like to implement a revised home page that could produce gains like this.
There are numerous detail “things” that were done by both the client and us to get to this stage. However, when I look back on the sequence these followed, I see some very simple principles that anyone can work through. Neither time nor space allows me to share them all; nevertheless, here’s five principles that, when applied, can deliver an unfair share of their responsibility for the probable success you will achieve.
Principle One – a Client with a Willingness to Test
We were fortunate to engage a client who valued their website not on how it looked but how it performed. They also had a technology partner who, whilst they were responsible for the existing website, were client focused enough to want to provide any assistance they could. So, when we floated the idea of split testing home page versions to improve the overall website’s performance, both the client and their technology partner were all for it. Without this approval, none of the gains from these changes would have been realised.
Principle Two – Knowing what you Don’t Know
I’ve written before on the power of knowing what you don’t know – here’s the prior article for those that missed it. This client wasn’t afraid to freely admit that there were some gaps in their understanding of the profile of their ideal prospect.
So we worked with them to bridge this gap by surveying their prospect audience with two very simple but nevertheless very powerful questions. (We have asked these same questions before to achieve the same goal for customers from a wide range of divergent industries from home cleaning to wardrobe manufacture to immigration consultancy.)
Principle Three – Quickly Applying this New-Found Knowledge
Once over 500 responses to these questions had been collated we went to work classifying and prioritising them. Then we re-wrote the key messages of their home page to better match the desires of their audience. We probably changed about 125 words of a 1250-word homepage. The images needed work too – but again not many, there were less than 10 that had some re-design efforts by Ben here.
Principle Four – High-Volume Pages Produce High-Speed Test Results
This client operates with an online sales process that spans four pages. As per usual the amount of traffic drops as visitors work through the stages. Nevertheless, the faster we could run each test the more we could fit into a month. So we applied the reworked copy and images of our initial tests to the first two stages of the sales process where the visitor count was at its highest. This reduced the cycle time between tests and helped us deliver improvements quickly.
Final Principle – Principle Five – Knowing there’s Always More to be Gained
The results shown in the image above were actually taken from the second test we ran on this page. The first gave a sizable improvement when compared with their original page – from memory the gains were similar again – but we could see there was more to be squeezed out.
On a review call, the client shared with us some new content that was ideal to use to create our second variation. There’s always more to achieve. As I’m writing this, one of the guys is working with the client on version 3, which we hope will push them through the 20% conversion barrier for this stage of their sales process.
So there you have it. Just five simple principles that you can apply when deciding which tactics to implement. These relatively simple steps, when followed, can produce some relatively sizable improvements.