Increase Conversions with Experiments & Split Testing
Split testing and experimentation is all about learning about your audience and how they want to interact with your business. By comparing just one element, you can learn what causes increases and decreases in audience engagement – making it a vital activity for increasing your business’ conversion rates.
Split testing (also referred to as A/B testing) is a method of conducting experiments with the goal of improving any metric, such as clicks, form completions, or purchases. As an experienced gambler would tell you, it’s important to hedge your bets, support more than one possible outcome, to ensure that your business does not end up betting an un-tested advert, landing page or email to deliver results.
Do you ever notice on a website that the content you see differs from what your colleague or friend may be experiencing? Sometimes these experiences are small, a heading change here, a different picture there, other times the page could be completely different. This is no accident, this is a deliberate experiment conducted to improve metrics, such as clicks, form completions, or purchases.
Which introduction did you prefer? Yes, that’s right, this is a split test in an article about split testing. In fact, there are a number of split tests happening on the page right now, so that we can determine which elements provide the best experience. From search advertising to email, we’re going to show you how you can split test too to increase enquiries and conversions.
There are a number of split testing opportunities available in AdWords to ensure that your business is delivering the optimum message to generate advert clicks and drive conversions. Within each Ad Group we recommend creating at least three adverts which each provide a different call to action or headline to understand fully which performs more successfully in fulfilling your business goals.
As you can see from our example, whilst our overall message remains the same, how we deliver it changes slightly with each advertisement so that we can understand what resonates best with those looking for SEO services. To create a level playing field, we recommend implementing the “Rotate Indefinitely” ad rotation option so that your adverts will be seen more evenly by users.
Whilst it can be easy to be trigger happy when it comes to making changes, it’s important that, when implementing a split test, there should only be one specific difference, otherwise it can be difficult to understand what worked, and what did not. If you were to change the call to action, headline and description, it can be difficult to discern what exactly caused the upturn in engagement.
You want to be able to say with confidence that by amending a specific element you saw an increase in conversions, otherwise, that way you can design future emails knowing that one call to action is more successful for your business than another.
Drafts and experiments let you propose and test changes to your Search and Display campaigns. They help you to measure your results to understand the impact of your changes before you apply them to a campaign. For example, you could trial two different landing pages within your campaign to understand which performs more successfully in driving enquiries and purchases on your website.
Here’s how to create an experiment in AdWords:
- Select your chosen campaign and then select Drafts in the top right corner of your screen.
- Select Create New and then name your Draft.
- Once submitted, you will be taken to your draft campaign, where you can make changes without affecting the existing campaign.
- After you have made all the necessary changes for your experiment, click Apply, and then select Run an Experiment.
- Name the experiment, enter a Start and End Date, and then select the Experiment Split. We recommend a 50/50 split to ensure that the experiment is shared with a similar audience size.
- Click Create, and your experiment is ready to go!
Conversion Rate Optimisation
So, you’ve created a landing page. You’ve followed all of the industry best practices to ensure that your page is fully optimised to generate conversions, however, there’s no one size fits all when it comes to landing page design.
What’s succeeded for one, won’t guarantee success for you. Instead, you should rely on trial and error to ensure that your page is engaging for potential customers. Some of the elements you should consider testing are:
- Call to action
- Page heading
- Page Layout
- Form length
- Copy length
- Main image
- Button design & colour
- Informal vs formal tone
There are a number of excellent split testing tools on the market, including those by Hubspot and Unbounce, however, the majority are pay to use tools. For a free A/B testing to help your business experiment with different variations of your website to deliver personalised experiences we recommend Google Optimize. Integrated with Google Analytics, Optimize is an easy to use tool, which provides in-depth insights to overall user behaviour on your website without spending a penny! However, it must be noted that there is only a three experiment limit at any one time.
Using Google Optimize, we’ve made some slight tweaks to this article to test which elements perform better than others.
An excellent way to gain repeat business, as well as new customers, email marketing provides an easy and cheap solution to communicate with an audience who have interacted with your business previously. Split testing can be an excellent tool to try new techniques or formats within your email campaigns – without the risk of sending an disengaging email to your entire data set.
Similarly to optimising your web page, there are a number of elements you could consider testing:
- Subject line
- Call to action
- How the reader is addressed
If you are looking to improve your conversion rate and engagement metrics, get in touch with one of our expert consultants today. We focus on objectives, segmentation, personalisation, deliverability and the all-important message to ensure you’ll achieve the most from your campaigns.