I was thinking about posting SEO new year resolutions today, but I think a lot of other blogs have already done that in a very good way, so let's tackle another very important (and sometimes neglected) element of a good online marketing strategy; website testing and particularly split testing.
What is website testing?
Website testing can take many forms, such as usability testing, UI/UX (user interface, user experience), technical tests or content tests such as A/B testing and multivariate testing.
Testing is a way to ensure that everything on your website is as it should be and that visitors can find what they're looking for.
While it should be logical for any business to do regular tests, assumptions and views of business employees are very often relied upon instead of looking at visitors and potential customers' opinions and needs.
Why implement split testing?
In difficult economic times, ecommerce businesses need to make sure they get the best ROI (return on investment) from their marketing and overall spend, and it is therefore crucial to ensure the highest level of conversion on your website. Many sites experience bottlenecks in their sales funnel on specific pages and changes in content, calls to action and offers can often have a dramatic impact on conversions.
Which methods to use?
I've recently been more exposed to the different tools available for testing, so here is a quick overview of the 2 main content testing methods:
- A/B testing - with this method 2 pages are used concurrently and displayed to different users in equal numbers. The 2 pages are matched with a common conversion point (payment confirmation page, sign up page....) and conversion results are compared to decide which page is performing better than the other. This method is very useful when testing new designs as it will very quickly highlight what works best for your visitors.
- Multivariate testing - this is a more complex tool which allows businesses to dynamically change predefined elements of a page in order to test their performance against conversion points. The changeable elements are predefined and will be displayed according to specific rules until the best performing combination is identified.
This method can be used to test content as well as graphic elements and can very often lead to dramatic improvements in conversions.
One thing to remember with these methods is that results will only be significant if you test the right pages, so identifying what to test plays a key part in the testing methodology.
Which tools to use?
There are a lot of tools available on the market, but if you're new to split testing, I'd advise to go with an agency at first to get a better understanding of how testing works, and how to set everything up.
For those of you who are more experienced, then Google Optimizer provides a very good structure to start testing as long as you have access to the html code of your website as testing requires the implementation of tracking tags within your site.
When looking at agencies, make sure to choose the right one for you as the most advanced solutions will be out of reach for most SMEs (from £100k +/year!). Some Google accredited providers will offer much more affordable testing services.
I am due to meet a few agencies to discuss split testing this week so I'll let you know what you can expect if you're considering testing yourself.
I think it's time to end this post and while I do realise that the content above only covers a small part of what can be tested on a website, I hope it helps you get a good overview of the basics.
Enjoy your Sunday.