You may think marketing is a fun, fluffy activity that anyone can do. At Redline we know that this is not strictly true; yes, we need to be creative and have the ability to think outside of the box, but it’s equally important to understand data driven processes and the role of analysis. Data is behind every decision, as it is the only way to get the best results from a campaign.
So, unless you have a crystal ball, in order to know what will work best, we need to test each campaign to collect the necessary data. Split testing is the simplest way to identify what works the best for your audience.
What is split testing?
The name is a give-away, it is a method of splitting and testing elements of a marketing campaign by running and comparing two versions simultaneously. The aim is to see which version performs better with your audience, by creating a ‘control’ and a ‘modified’ version (let’s call them A and B). Version A would be shown to one group and version B to the other, and the results measured against pre-decided metrics, for example, this could be the click through rate or the conversion rate.
What are the benefits of split testing?
Split testing is great for digital marketing, you could test individual elements on a webpage to see the impact on the functionality and user behaviour. It is particularly effective for optimising landing pages, as the smallest change can make a dramatic difference to the conversion rates. Similarly, with email marketing campaigns, split testing can help to improve engagement, this could be as little as changing the subject or the layout.
-
Increase conversion rates
A small change, such as the colour, text or position of a call-to-action button can make a dramatic difference, you will only know by split testing.
-
More relevant content
If you know what your audience likes, you can create more of the good stuff and less of what they don’t like.
-
Improve user experience and increase engagement
If you know which version is more popular, you can do more of the same to improve their user experience and satisfaction.
-
Cost savings
Minimise risk by making changes that aren’t data driven, allowing for effective and efficient resource allocation.
How to split test
The premise is simple but there are eight rules that should be followed in order to get the data you need to make informed decisions.
- Choose your goal: To test and measure effectively, you need a clear goal. This could be increasing click through rate or open rates for an email.
- Choose your variable: To establish which tweak works, you need to only implement one variable at a time, for example the colour of the call-to-action button or the position or the size, but not all three together.
- Create your assets: You need two versions of the same asset; one is the original ‘control’ and the other the variable featuring the specific change you are testing.
- Split the audience: Split your audience into two equal groups (randomly) to make sure they are similar; send one group the control and the other the variable version.
- Observations: Choose your split testing tool such as Google Optimize to help run your analysis.
- Simultaneous campaigns: Run both versions of the campaign simultaneously until you have enough data.
- Analysis: Analyse the results to see which version the audience preferred.
- Data driven decisions: Implement any necessary changes and apply insights to future campaigns.
If we did have that crystal ball… we would see that your future holds, testing, testing and more testing!
