Welcome to Overthrow Digital's blog where we share our thoughts and ramblings on anything and everything to do with the world of digital communications.
17
Feb
2015
Written by:
David Slade

A/B Testing your website and campaigns - Part 2

Part 1 of this blog post discussed what A/B Testing is and what the most common variables that people test to see results on their website and digital marketing campaigns. Part 2 will look at areas to test outside of colour and images.

One thing to bear in mind is that an increased conversion rate doesn't always result in increased sales, but it can help...

 

Navigation Menu's 

 

Menu's can clutter pages and aren't always necessary for every page. A/B Testing results have often shown improved click through's and better use of space when the navigation menu has been removed for specific pages.

 

Modifying layout

 

You would be surprised how much of a difference the layout makes and the impact it can have on your results. This can also have great effect with questionnaire fields. Instead of having a long list of 10 questions beneath each other (in a portrait fashion), experiment with a landscape field with 5 questions on each side. This makes the questionnaire form look much shorter although it still has the same amount of questions. 

 

Line spacing 

 

Increasing the line spacing has produced good results. The effect is that the wording looks much less cluttered and condensed which makes the page look much more presentable. More importantly, increasing the line spacing means that you can't fit as much text on the page (less reading is always more appealing).

 

Questions

 

Adding, refining or removing questions can have an impact on the results too. Questions feel personal and require direct interaction plus an individuals opinion (every one likes to add their two pence) You can also use the answers as a source of info to generate insights as to what might improve results. (or what isn't working in your favour).

 

Headlines 

 

With the headline typically being the first thing a user will read, you can either draw them in or put them off with the headline. Pay close attention to wording here. A good idea is to think of what you would want the headline to say if you were looking for the service you are offering. What would draw you in?

Combining these techniques as well as the more common variables covered in Part 1, with the Google Analytics results, you can start identifying people's patterns and behaviour on your website. Seeing where on your webpage people are clicking can highlight sweet spots that you might not be aware of. Once aware of these types of statistics, you can re-arrange your layout and move your call to actions to these locations.

 

Is it possible to run A/B tests on things other than web pages?

 

Yes. As well as landing pages, many people run A/B tests on CTA's, Email campaigns & PPC.

Call To Action's - With CTA's, try changing the messaging on the CTA, it's shape and size, colour or placement on the page.

Email - Email testing variables include the subject line, personalisation features, sender name etc.

PPC - For paid search ad campaigns, you can A/B test the headline, body text, link text or the keywords. Something worth considering is the appearance of your website on mobile and tablet. Mobile and tablet internet browsing have now over taken desktop usage, which means that it's more important to be concerned with how your site is performing on mobiles and tablets more than ever before.

Some of the most reputable split-testing software platforms are listed below:

(This info is from the conversion rate experts website)

- Accenture

- Adobe

- Autonomy Optimist

- Avenseo

- Google Content Experiments - Free 

- ClickThroo

- Convert

- CRE Engine

- GetSmartContent

Overthrow Digital 7 Albert Mews Albert Road London N4 3RD