Skip to main content
A/B testing for healthcare

A/B testing for healthcare: What the top organizations are doing

July 13, 2021
Reading time: 
Daniel Boltinsky
Daniel Boltinsky
Kameleoon, Managing Editor, North America

Counterintuitively, better technology does not seem to be the X factor for industry leaders in digital health. Instead, the difference-maker for top players in healthcare, according to the digital healthcare researcher and designer Hanno, is patient experience.  Recent research by Forrester, which you can see here, reflects this: They found that healthcare organizations (HCOs) that were able to “discover and apply customer insights are five times more likely to grow [their] revenue.” 

One of the most effective ways for your HCO to harness patient insights is with A/B testing (also called split testing). A/B testing collects data on how users interact with two different variants of a webpage, app, or anything else.

By comparing how users interact with two different experiences, you gain insight into how your technology’s user interface affects user behavior — insight that helps you optimize your website based on what people genuinely prefer, not just what you think they need.

HCOs looking to stand out in the rapidly growing digital healthcare industry need to start putting their patients at the heart of how they optimize their products and website. A/B testing can increase patient satisfaction through patient-centric, data-backed improvements to your website’s accessibility, usability, and personalization.

Download the report to see how HCOs are using data insights to improve patient experience.

healthcare report

1 They use segmentation and data to drive personalization

Segmentation allows you to divide up your patients based on shared characteristics, so you can test individual groups, understand what’s relevant to them, and personalize their experiences. Without segmentation, it is nearly impossible to find relevant insights in your data because each patient segment may have unique or conflicting needs.

For example, a webinar on child development might not get great sitewide engagement, but it could be hugely popular in your segment of women ages 25 to 35. Isolating this segment in your A/B tests can show you how to target that audience better. For instance, you could A/B test the copy in your email marketing campaigns and landing pages to optimize for the pain points that matter most to women in this cohort, leading to higher email open rates and sign-ups for your webinar.

Segments are only helpful if they’re meaningful, cohesive groups. A customer data platform (CPD) can help you pick apart your data to create the most relevant groups on your website. When you’re A/B testing using intelligently created segments, you can:

  • Find pain points for individual groups and use them in your marketing outreach.
  • Tailor landing pages to offer relevant content to that segment.
  • Design web pages customized for a specific location.


By dividing your patients into segments, you can clearly see what each group is looking for when interacting with your product. With this data, you can tinker with your messaging, design, and offerings to better fulfill your patients’ needs, giving them the personalized experience people want.

2 They challenge existing ideas in the pursuit of better patient experience

America’s healthcare systems can often be unfathomably complicated and unhelpful; sometimes, even experts struggle to get the care they need. A/B testing helps you overhaul these needlessly complex systems, but only if companies are willing to experiment and occasionally fail in pursuit of an improved product. By fostering a culture of experimentation when it comes to user interface (UI) improvements, you can find what works best for all of your patients, no matter who they are or what healthcare needs they have.

Not every A/B test you run is going to be a winner. In fact, most will fail or lead to inconclusive results, but that doesn't mean what you learn by doing these tests isn't useful. For example, if your HCO is targeting millennials, a group looking for a convenient way to fit doctors into their busy lifestyles, you may want to test out different ways to help improve the convenience of your product. This could mean:

  • Testing shorter appointment times
  • Offering appointment confirmations by text or email
  • Suggesting targeted articles that can teach patients how to improve their health at home


Each of these is a hypothesis about what you believe would be convenient for millennials accessing your product. Even if only one of them actually works, you’ve already improved your patient experience a little bit.

For the failures, go back to the drawing board and ask yourself whether you implemented the test correctly, if the hypothesis was wrong, and what these results tell you about your target market. The more you hypothesize, test, and adjust, the more your product will appeal to people fed up with or locked out of traditional medicine.


user consent management

3 They balance patient experience with HIPAA compliance

One of the largest hurdles to personalization in the healthcare field is a fear of compromising patient privacy. Many HCOs worry that digital customer data, especially third-party data or pseudo-anonymized customer data, is risky to keep or optimize with.

In the U.S., the primary law HCOs need to be concerned with is the Health Insurance Portability and Accountability Act (HIPAA). HIPAA charges healthcare providers with safeguarding their patients’ sensitive data, including SaaS software that might handle protected health information (PHI) while running A/B tests.

However, worries about HIPAA are often overblown as a lot of personalization efforts can be done without worrying about HIPAA at all. Using anonymous behavioral data like bounce rate, time on page, and conversion rates, you can set up A/B tests that improve user experience (UX) without ever involving PII.

For example, you could hypothesize that your current pricing page is too complicated. You could then A/B test a second simplified pricing page and compare it against the original’s conversion rates. With those results, you can give patients a more relevant, optimized pricing page without risking their PII.

Nevertheless, you will want to take the proper steps and use the right tools in order to expand your A/B testing program. Indeed, once you’ve gained maturity and quick wins, you will need to incorporate PII into your strategy while staying completely compliant with HIPAA.

A few basic measures include:

  • Sharing only personal data needed for your desired patient experience improvements
  • Asking how SaaS platforms comply with HIPAA on a technical level (i.e., secure data transfer)
  • Signing a Business Associate Agreement before handing over any data

4 They leverage software for advanced testing capabilities

A/B testing has come a long way from being a simple test between two variants of a webpage or email.

Although this kind of basic A/B test is still effective, there are many ways to boost testing performance using new, advanced software. Taking advantage of this new tech can not only speed up your testing process but can also help you get more detailed insights from the data you collect from your patients.

Here are just a few advanced testing features you could use to boost your A/B testing:

  • Multi-Armed Bandit Testing: Also called dynamic traffic allocation, multi-armed bandit testing software combines the testing and implementation phases of traditional A/B testing. As one variant shows it is superior, the software will send more patients to that winning variant, so you don’t miss out on those conversions.
  • Live Simulations: A live simulation allows you to see that your new variant works before it goes live. This way, you see for yourself if your variant works the way you had planned across different devices and personalization profiles.
  • Full-Stack Integration: When testing and customization live on separate platforms, implementing changes takes a lot longer. Full-stack integration means that your A/B testing software works with your other platforms to make changes seamlessly, saving you time and effort when testing.
  • Restrictionless Testing: Some A/B software puts restrictions on the number of variants you can test or the amount of traffic you can run those tests on. These limitations can slow you down or affect the results of your tests, so investing in a platform with restrictionless testing can help you better improve your patient experience.


If you want to take your testing to the next level, advanced A/B testing software is the best way to make that happen. By investing in top-notch testing platforms, you can improve your program’s speed, accuracy, and ease—delivering a data-driven personalized experience to your patients.

Put the Patient Back at the Center of Your Website

If healthcare organizations want to foster digital adoption and encourage increased patient numbers, they need to start using technology to make their products or services more convenient and accessible. To learn more about A/B testing, sign up for our free courses and certifications, where you can learn how to securely improve and personalize your patients’ digital healthcare experiences.


Download the report to see how HCOs are using data insights to improve patient experience.

healthcare report
Topics covered by this article
Web Experimentation
Daniel Boltinsky
Daniel Boltinsky
Kameleoon, Managing Editor, North America