There is a lot of literature available on the topic of website optimisation and driving traffic to your website. Of course, having visitors is the cornerstone of your online success, but the end goal is for them to become customers through buying products or signing deals. This is where a lot of business owners are left scratching their heads, as visitors to their site just don’t seem to be converting. In our experience, most of the time this is due to website usability issues and is somewhat of a human psychology science which we at This Side Up find quite fascinating.
This Side Up has the capability to redesign website pages, layout and navigation and run tests to determine their performance in creating sales or leads. In these tests, like with any experiment, you need to approach scientifically to ensure legitimate conclusions can be drawn. But how to go about it? Well, read our simple tips.
Set Your Strategy
Although conversion rate optimisation (CRO) can seem a bit haphazard as it relies on testing and elimination, it is nevertheless essential that you have a clear strategy in place to execute it well and get positive results. So before throwing yourself into A/B testing, spend some time on the following steps:
- Identify sticking points
What positions in the online sales pathway are contributing the most towards a drop in conversions? Perhaps it is that visitors exit your website on a specific page or point in the checkout process, or maybe leave shortly after landing on the homepage. It may highlight the fact that the journey you want them to take isn’t clear on that page, that something really puts them off; or that the page is not designed around what the visitor expects. These points are the low hanging fruit which if tackled correctly can give quick gains towards an increased conversion rate.
This second step will be very much dictated by what the first step has revealed. Depending on the issue, for inspiration to your required changes you may need to look at what the competition or other top websites are doing, study your customer’s online pathway to learn about their expectations and even go as far as conducting supervised studies of users navigating your website.
- Form a theory
From here you should then be in a position to form and define a theory about what particular features, if changed, would result in greater online conversions. For example “It is proposed that shifting the Buy Now button from the bottom of the page to the top will increase in a greater click through rate to the payments page” or “Adding a hero image of a customer using the product on the homepage will reduce bounce rates”.
- Develop the experimental method
Now you must consider how the experiment will be conducted. Will it be a simple A/B test where 50% of visitors are sent to variant one and the other 50% to variant two? Or is it a more complex, multivariate test where combinations of changes are tested on smaller segments of traffic? Make sure you cover all bases of your test. It is advised that only small and individual changes are implemented in order to isolate these variables in your tests for better understanding of their contribution to conversions, however in some cases this is hard to avoid.
- Run the test and analyse the results
This is, of course, one of the most important stages of the process. Tests are of no value without metrics to highlight their performance. Remember, this is science! One thing to keep in mind is statistical significance to ensure meaningful conclusions but if you can’t remember what your boring highschool stats teacher was on about don’t worry just have a quick read over this guide.Test, test, test! Believe it or not but your site is never perfect and so always adopt an attitude of continuous improvement. User expectations change over time so there are always tweaks to be made.
If you are interested in increasing your website’s conversion rate, get in touch with the split testing experts at This Side Up to take the first steps. Or for more tips on where to start for improving your website conversion rate, have a read of this recent article we wrote.