Why you should split test?
Well it is vital that you understand what motivates your visitors when you’re marketing to them you need to know what turns them on
You have to Maximise conversion for your campaigns. Everything you do costs money in someway whether its time with indirect costs, or direct costs of collateral.
You have to deliver efficient advertising. The most you can get out of every campaign is going to be an increase to your bottom line.
What should you split test?
Well each step of a campaign is designed to achieve a goal, and it’s the success of each of these goals that needs to be measured and tested.
In the example of an email campaign there are some distinct elements that each have a purpose;
The email subject has the job to stimulating sufficient intrigue or interest so the recipient opens and reads the email.
The first line of the body needs to build on that to maintain the interest, and the remaining body text needs to compel the reader to take action, for example click through to a landing page.
The landing page then has the express role of ensuring the visitor now takes the required action perhaps that’s to sign up to a mailing list, buy a widget or register their interest in a service, in other words they somehow make a connection back to you.
Each of these steps should be tested to see what approach gets better results. Which subject line got more opens, which body content got more click through.
When you should split test?
Well certainly for paid advertising, particularly if you’re doing any kind of online advertising with AdWords or other PPC or PPM advertising.
If you’re using campaign landing pages, have two different ones and see which approach generates the most sign ups, you can even try something wacky to see what impact that has.
Any new web design your producing, whether is a home page, or a complete re-skin, or just specific pages, or in fact ay page with a required outcome should be tested to understand the most effective approach.
You can even split test you remarketing campaigns, which I was discussing in last weeks email.
So, what tools are available to measure results?
Many of the established tools come with built-in support for split testing, for example if you use Google AdWords for PPC campaigns that provide the option to create different ads per ad group and it then automatically starts using the one that is the most effective.
Google Analytics and in particular Google Analytic Experiments allow you to manage a variety of split tests and can even manage certain aspects of content on your pages to control who sees what version, which button style or even the colour scheme.
If you’re a WordPress user then there is a good selection of plugins you can choose from, one popular one is called Neilo
And there are many other tools developed by many independent developers, often to cater for niche markets that you can use.
What should your next steps be?
You need to start learning the behaviour of your own customers and your own industry.
Pick a campaign and try to use a large an audience as possible to make the results as meaningful as possible; when analysing percentages you get better results from a larger data source.
Select which one thing you want to test – don’t test two things at the same time as you won’t ultimately know which one thing made the campaign successful or unsuccessful.
Run the campaign the watch to see what happens and make the changes to the next campaign. In the example of an email campaign, once you understand the approach to the subject line, move on and test the body, but never stop testing.
Go back and re-challenge your understanding and make sure there isn’t now a better approach then there was 3 months ago to make sure you’re always fine tuning every campaign to be as successful as it can be.
If you have any questions about this, or you need any more advice or information please don’t hesitate to contact us by completing the form on this page and we will of course get in touch with you to see how we can help.