Split testing, also known as A/B testing, is a relatively simple and cost-effective way to experiment with different types of content and creative.
What you look for in split testing is a pattern of behaviour from your audience on what sort of content or copy they prefer over another, in other words, you’re testing Item A against Item B.
While there are many things that can be examined in split testing (e.g. timing, location), we’re going to take a look at split testing creative content.
Why should I do it?
A/B testing is great for revealing information about a number of topics:
- If your ad click-through-rates are slowing
- If customers are bouncing off your website
- If customer conversion rates have dropped
- If you’re looking to go into new markets
- If you’re launching new products and services
How do you start?
First, you need to have an understanding of your current click-through rate. This will become your campaign benchmark and help you see impartially whether your testing offers successful results.
Second, you need a sample audience; this can be in the format of a subscriber list, customer list or even an existing audience on one of your online advertising channels (e.g. industry listings, search engines, social channels).
Third, you need at least two versions of the item you want to test. If you’re testing elements of creative, then you’re testing copy, design or imagery used.
Then you need good tracking software that can count up your click-throughs and trace customer behaviour.
What do you choose to test?
This can be as broad as you want to make it. Creative-wise, you’ll be looking to fine-tune the way you show your brand and communicate your message to your customers. This could be through:
- Increasing design elements (e.g. enhancements, effects and colour)
- Decreasing deign elements
- Changing the language
- Reducing the number of words you use
- Increasing the copy descriptors
- Focusing on image-based calls-to-action (e.g. progress shots, finished products)
What should stay the same?
Everything else! To get a true reading of customer feedback, you want to run your A/B ads concurrently at a normal schedule so they can be tested against the same sample audience at the same time. Theoretically, it means that an unbiased sample is seeing one of the two ads and responding naturally. Plus, if you have real-time feedback, you’ll be able to see which advertisement is gaining ground straight off the starting block.
What will I be able to find out?
Your split testing should focus on addressing one or two issues at a time, but in general, your data should be able help pinpoint whether:
- There’s a strong creative element that’s most attractive to your customer
- The creative aligns with your website messaging
- Your website is letting your ads down
- The right audience is drawn by your creative
- There’s another element that needs tweaking, and more
Sometimes we see that creative A/B testing doesn’t impact the click-throughs – we remain at the same benchmark figure or worse. This sort of result indicates that there are perhaps other elements in play that need addressing – this can be any combination of product, promotion, price, placement, and timing issues.
Can I do this on my own?
Yes, you can, but you do need to be prepared to knuckle down with some clear objectives, follow-through and results measurement. A/B testing is best conducted on the same days over several weeks to build a meaty database of results – so it requires proper dedication.
That doesn’t mean you have to go at it alone, marketers and marketing agencies like ours help hundreds of businesses test and refine their ads to create campaigns that truly “speak” to their audience.
If you would like more information on split testing and how it can help your business reach its goals sooner, why not give us a call today!