A/B testing in HubSpot lets you compare two versions of an email, landing page, or CTA to see which one performs better with your actual audience. You run both versions simultaneously, split traffic or recipients between them, and let the results determine the winner. The goal is to make decisions based on data rather than assumptions.
Pick one element to change — subject line, headline, CTA text, or send time. Testing multiple changes at once makes it impossible to know which change drove the result.
For emails, open the email editor and click “A/B Test” in the top toolbar. For landing pages, open the editor and select “Run A/B Test” from the page settings.
HubSpot defaults to a 50/50 split for most tests. For emails with large lists, you can test with a subset (e.g. 20% each) and send the winner to the remaining 60% automatically.
Choose what counts as winning — open rate, click rate, or conversions. For sales-focused campaigns, click-to-submission rate is usually the most meaningful metric.
HubSpot shows a confidence score on the results page. Aim for 95% confidence before declaring a winner. Running a test for too short a period gives unreliable results.
What Can You A/B Test in HubSpot?
HubSpot supports A/B testing across several tools depending on your subscription tier:
- Marketing emails — subject lines, preview text, sender name, email body, send time. Available on Marketing Hub Starter and above.
- Landing pages — headlines, hero images, form placement, CTA copy, page layout. Available on Marketing Hub Professional and above.
- CTAs (calls to action) — button text, colour, placement. Available via the HubSpot CTA tool across Professional tiers.
- Website pages — page copy, layout, trust signals. Requires Content Hub access.
If you are on a Starter plan, email A/B testing is the most accessible starting point. Landing page testing requires at least a Professional-tier Marketing Hub subscription.
How to Set Up an A/B Test on Emails
Navigate to Marketing > Emails in your HubSpot portal. Create a new marketing email and open it in the editor. At the top of the editor, you will see an “A/B Test” button — click it and HubSpot will duplicate the email, creating a Version A and Version B.
Make one change to Version B. The most commonly tested element is the subject line, because it has a direct impact on open rates and is easy to isolate. Once both versions are ready, set the split percentage and the winning metric (open rate is the default; you can change this to click rate). Schedule or send, and HubSpot will automatically track results per version.
If you enable automatic winner selection, HubSpot will send the winning version to the remaining contacts on your list after a period you specify — typically a few hours after the initial send.
How to Set Up an A/B Test on Landing Pages
Go to Marketing > Landing Pages, open the page you want to test, and select “Run A/B Test” from the Actions dropdown. HubSpot creates a duplicate of the page as your Version B.
Edit Version B to reflect the change you want to test — a different headline, a shorter form, a different CTA position. Both versions will receive live traffic according to the split you set. HubSpot tracks views, submissions, and the overall conversion rate for each version in real time on the landing page details panel.
Unlike email tests, landing page A/B tests run continuously until you manually end them. Check results weekly and end the test once you have enough traffic (typically a few hundred views per version) to reach a reliable confidence level.
How to Read A/B Test Results
HubSpot shows the results for each version side by side on the email or page details screen. For emails, you will see open rate, click rate, click-to-open rate, and unsubscribes per version. For landing pages, you will see views, submissions, and conversion rate.
The most useful figure is the confidence percentage, which estimates how likely it is that the observed difference is real rather than random variation. A confidence level of 95% or higher is generally considered reliable. If confidence is low, let the test run longer — ending it early on the basis of early data tends to produce false conclusions.
Combine A/B results with your HubSpot reporting dashboards to track how test winners affect downstream metrics like deal creation rate or pipeline contribution. A higher click rate on a test does not always mean more revenue — checking the full funnel matters.
A/B Testing Best Practices for Southeast Asian Teams
- Test one thing at a time. If you change the subject line and the send time in the same test, you will not know which change produced the result.
- Run tests long enough. For emails, a few hours is usually enough once the list size is above a few thousand. For landing pages, aim for at least a week to account for weekday and weekend traffic differences.
- Segment before testing. A test run across your entire contact list may mask differences between segments. Use HubSpot lists to run separate tests for different audience groups — for example, Thai-language contacts versus English-language contacts, which is a common split for teams across Southeast Asia.
- Document your results. Keep a simple log of what you tested, the result, and the confidence level. Over time this becomes a useful reference for what works with your specific audience.
- Do not stop at one test. The winning version of a test becomes the new baseline. Run the next test against that winner, and so on. Each round narrows the gap between your current performance and your ceiling.
What to Do After an A/B Test
Once a winner is declared, update your default email or page to use the winning version. If HubSpot’s automatic winner feature is enabled for emails, this happens automatically. For landing pages, you need to manually end the test and select the winner.
Use the result to inform your automation workflows. If a shorter subject line consistently outperforms longer ones, apply that learning across your enrolled nurture sequences. If a particular CTA text converts better on landing pages, use it across related pages in the same campaign cluster.
Frequently Asked Questions
Do I need a paid HubSpot plan to run A/B tests?
Email A/B testing is available from Marketing Hub Starter. Landing page A/B testing requires Marketing Hub Professional or above.
How long should I run an email A/B test?
For most marketing email tests, a few hours is sufficient once the majority of the list has received the email. If your list is smaller (under 1,000 contacts), results may need more time to stabilise.
Can I A/B test HubSpot workflows?
Workflow branching (conditional logic) is not the same as A/B testing, but HubSpot’s Professional and Enterprise tiers allow you to split workflow paths and track downstream outcomes per branch — a useful proxy for testing sequences.
What is a good sample size for an A/B test?
As a general guideline, a few hundred responses per variant is enough to reach a meaningful confidence level for most email metrics. For landing pages, aim for at least 200-300 visitors per version before drawing conclusions.
Where can I find the A/B test results in HubSpot?
For emails, click on the sent email in Marketing > Emails and select the “Details” tab. For landing pages, open the page and check the “Performance” tab. Both show version-level breakdowns.
This is an independent informational resource produced by a HubSpot Platinum Partner. HubSpot is a trademark of HubSpot, Inc.
Planning a HubSpot setup or integration in the region?
Contact us