If your business has a significant online presence, such as a website or an app, the chances are it could benefit from some A / B testing.

By Eoin O'Hara

Ah, but what is A / B testing, and how exactly might it benefit your business?  

The process of A/B testing can be used in all manner of different fields, but for our purposes, it applies to the action of comparing two different versions of the same web page with the intent of discovering which of them performs best in achieving a goal or goals.  

It might sound fiddly and complicated but it isn’t, and in the case of most businesses it simply tests which version of a webpage delivers the highest conversion rate i.e. the one which is best in encouraging visitors to take the desired action, such as clicking through to a product description, service, or other source of information.    

In its most basic form, A / B testing can be carried out by simply observing friends, family, or co-workers whilst they navigate your site, but of course there are many more sophisticated and scientific ways of doing it, which yield quantifiable and actionable results.    

Nowadays A / B testing is available right at your fingertips, with countless online services offering to deliver comprehensive results fast. Traditionally these work by automatically showing two variants of the same webpage to visitors at the same time. The software will then analyse how each visitor behaved whilst on the page, and use this analysis to inform its decision on which version encouraged the desired response.    

Repeat this process over and over again, and you will soon build up an accurate set of data which will hopefully allow you to implement positive and informed changes.    

But how can a bootstrapped startup possibly afford such a sophisticated and no-doubt pricey system?

Well, the good news is, most services work on a freemium business model, meaning that for many startups with a low to medium level of site traffic / complexity there are some great free options to be had*.  What’s more, most modern services don’t require any coding experience or specialist technical skill, which makes the process significantly more accessible for entrepreneurs than it was just a few years ago.   We have done some cyber digging and found a countless number of services aimed at the non-technical entrepreneur all offering greatly varying levels of testing complexity.    

Generally these fall into three levels of ‘sophistication’- we have selected one example of each of these levels below that should give you a solid starting point for your A/B testing, without having to spend a penny.    

* Beware, as most services offer a free trial which converts into a monthly contract if not cancelled within a set period of time.  

Level 1: 5 Second Test  

This software gives your users an active role in the A / B testing of landing pages and calls to actions by assessing how much of your design and placement of prominent material they remember within just five seconds.    

The idea is that you can gain substantial insights into how users perceive your webpages and the most important information, which you display on them. Over time amendments and revisions can help you to hone in on the version which provides the best feedback in relation to your goals, and eliminate distracting / irrelevant material which might be serving as a barrier to growth.     

There is a wide variety of test formats available to users including a click tests, and preferences test, and whilst we haven't tried out all of them yet, so far it seems fairly straightforward and easy to implement.  The most basic products are free as well with the option to ‘go pro’ so there’s no risk in giving it a shot.   If you are new to A / B testing, this might be a good place to start.    

Level 2: Google Analytics Experiments  

For those of you who are a little more confident in your A / B testing acumen, some time spent using Google Analytics can really help to turn some of your insights into actionable points for your apps and web pages.    

It allows you to use your existing analytics account to create up to ten experiments on any individual page, each with its own variant to be tested. What’s nice about this system is that it gives you significant control over the conditions of your experiments, as well as the amount and specificity of data that you receive.  For example you can easily set the duration of the testing process, and specify the percentage of users who are to be included in the study, meaning that potential negative impacts on your current operations can be managed quite simply.  

Once created, you can be updated on the progress of your experiments via email, allowing you to get on with other things without having to worry too much about progress or effectiveness.   Like most Google products for webmasters and developers, it’s quite easy to use, which is always a good thing!  

Level 3: Wingify and their Visual Website Optimiser  

You may well have heard of Wingify in the media recently; they are an India-based startup and have made big strides in increasing the accessibility of A / B testing to the average non-technically minded entrepreneur and website owner.  

Their Visual Website Optimiser (VWO) combines some of the traditional aspects of A / B testing with emerging technologies which give less data-heavy, and more easy to interpret results.  

In addition to the standard A/B testing facilities offered by the VWO, the software also includes some other tools which compliment this service allowing you to more actively engage with the results and implement changes accordingly, these include heat maps, user insights, and user personalisation capabilities.   Heat maps, for example are growing fast in popularity as a way of raising conversion rates by assessing where your users are focussing their attention whilst on a particular page. The information collected during a visitor’s time on your page is used to generate easy to interpret ‘heat maps’ which instantly allow you to identify how effective your important messages or calls to action are at grabbing a visitors attention.   You can then compare and analyse how individual alterations to a page impact on a visitors behaviour, giving you a broad indication of whether an improvement has been made.  

What's also quite nice about this service is that it gives you the capability to personalise user experiences based on previous behaviour, and thus present them with information (such as deals) which are highly targeted and specific to their individual preferences.  

Wingify is a good example of a more comprehensive testing and UX analysis service. It offers a 30-day trial of its services after which you will roll over into a monthly contract unless you cancel it, so be careful!