Data drives decision-making. But sometimes, data can be skewed or incomplete.
Even when we think we’re making the best business or marketing decisions, there are variables we might not be accounting for in testing.
One way to improve data collection inside your CRM is to test multiple variables simultaneously using split testing.
HubSpot’s A/B testing enables you to test out multiple versions of a variable, including emails, landing pages, CTAs, and web pages, to see which version drives the best results. Leveraging these results, you can improve upon existing steps in your deal pipeline to derive the most money out of your marketing efforts.
In this guide, we’ll explore how to use HubSpot’s A/B testing to drive the best results to your website, including the benefits of A/B testing, how to set up A/B testing, establishing the suitable test duration, and showcasing how to test each variable inside HubSpot.
Navigate to the following pages inside your HubSpot Marketing Hub to get started with HubSpot’s split testing program:
Marketing > Email > Create Email > Create A/B Test > Enter name of email B > Ok > Click Variation A or B > select variations (ex. subject line, etc.) > set distribution scale > select winning metric > set duration
Test results are published in the Performance tab and can be viewed during active testing or at the end of the duration.
Email A/B testing features also include:
Conducting A/B tests for web pages, landing pages, and CTAs is very similar and will require you to navigate to those pages in the HubSpot Marketing Hub and click on “Run a Test” in the dropdown menu.
The most popular variable to split test inside HubSpot is email. You can run tests on emails for standard drip campaigns or HubSpot Sequences.
HubSpot’s email A/B testing allows you to publish and track experiments using different:
• Subject lines
• Promotional offers
• Body copy
• Images
• Email senders
A/B testing emails allow you to examine which marketing material produces the highest email engagement via:
• Open rates
• Click rates
• Click-through-rates
• Bounce rates
Users can also optimize the distribution of each email, which lists they target, and the duration of each test before results are published, and a winner is selected.
Similarly, A/B testing landing pages allow you to deliver the best marketing ads for customers to improve click-through rates on your website and capture more leads.
• Experimental variables include:
• Page copy
• Offers
• Headlines
• Form fields
• Images
When you go to run an A/B test on a landing page, you may also see an option to run an adaptive test.
Adaptive tests are even more complex and allow you to run up to five landing page tests simultaneously using various variables. As a result, adaptive testing is best reserved for advanced marketers and can be more confusing than rewarding if not done properly.
A/B testing CTAs allows you to improve your user experience by creating a call-to-action that entices user action. In turn, improved calls-to-action can help you improve your conversion ratio, thus leading to higher lead counts and purchases on your website.
A/B test variations include:
• Different button sizes
• Different button text
• Different button colors
Finally, important web pages, such as homepages, checkout pages, and shopping carts, can be A/B tested to improve your conversion and dwell rate.
Testing variables depend on the page you’re testing but could include similar variations, such as images, headers, form fields, etc.Examine your analytics to see where improvements could be made to your campaigns. For example, are your drip emails receiving high open rates and low conversions? Then you might want to test out new body copies or CTAs.
Understanding where improvements can be made and setting clear goals to improve those metrics will allow you to focus on the variables that drive the biggest impact on your campaign.
One of the most common mistakes we see marketers make is failing to use a control variable. For example, suppose you have active emails running and decide to run two different variations of the same page but change the copy of one page and images on another. How will you know which factor improved your conversion rate if the results are similar?
Always run a test beside a control variable to acquire incremental results that improve marketing performance.
Furthermore, run tests on one variable simultaneously to gain precise results. For example, if you test out a new email headline, body copy, and featured image in a drip campaign, how will you know which variable contributed most to your click-through rate?
Test one variable during each test to gather precise results and cut down on guessing games.
Different audiences and contact lists will produce different results from campaigns. For example, some audiences may respond better to emails that invoke sales promotions in their headlines. In contrast, others may respond better to headlines that prescribe tips or information about a particular product.
Consider running tests for different audience segments inside HubSpot to personalize campaigns, so you achieve the best results.
It’s crucial to treat split testing as a long-term project. Therefore, we recommend only running one test per campaign at a time–at least while you get started.
Again, testing out different emails, landing pages, and CTAs at once can obfuscate results, providing no actionable results that can be gained from tests.
Allow time for tests to collect data to make more informed decisions. This tip depends on which variable you’re testing. For example, we recommend running A/B tests on shopping cart pages for up to a month or 30 days to gather the most accurate results. Tests could be extended even further for emails and Sequences based on how long it takes the average lead to convert.
On the other hand, landing page tests on ads that require continuous marketing spending may benefit from tests that only last a week so you can make quick decisions that reduce costs and spending.
Overall, it’s essential to customize test durations for each variable based on your business model and sales pipeline.
Finally, you can improve test results by collecting feedback directly from users. Collecting user feedback helps you learn why certain variables resonate with customers over others and allow you to optimize campaigns to reflect that sentiment.
HubSpot’s A/B testing is a valuable tool that can help improve data collection and provide actionable strategies. For help configuring A/B tests and your HubSpot data ecosystem, contact the experts at Coastal Consulting.
We offer HubSpot audits and consulting services to help you manage your HubSpot account, including your integrations. Contact us for your consultation and to take back control of your data today!