Remember those science classes, where you took a control group and tested it against an experimental group that had been changed in some way to see whether the change was beneficial?
Well, that's A/B testing: A basic science experiment redesigned to work as a marketing tool.
As with those experiments, A/B testing that's done well is an important source of concrete data. It supports or refutes ideas you might have about increasing your inbound traffic (your clients) so as to make your online marketing efforts more successful.
Although you're familiar with the scientific concept, let's look more closely at the marketing practice.
Here are the steps to an A/B test:
- Decide what you want to test, whether a single element, like a new call-to-action form or an image added to your Twitter posts—or multiple parts of an element, like a redesign of a website or landing page. (In other words, test just one element, regardless of how many pieces to that element there may be. An A/B test is not a test of combinations, like multivariate testing.)
- Devise an experiment changing the variable you want to test against a control group, ideally your existing call-to-action form, Twitter look, or landing page. The control group is Group A, and the variant it is tested against is Group B.
- Define the metrics for the test, or your exact goals in running the test, such as improving your click-on rates, conversion rates, downloads, etc.
- Split your viewers into two groups, with one being sent to your variable and the other to your control group.
- Create your A/B test and continue running it for a long enough period of time to collect statistically significant results, be willing to repeat the test to confirm your results.
That next-to-last point is a challenge for non-statisticians, and reason enough to probably want a professional to do the testing for you, even at a cost. But if you have a mathematically gifted employee in your practice, here's a guide to understanding statistical significance from Hubspot that may be of help.
Why conduct an A/B test in the first place? How might it help bring new clients to your practice, or better engage existing ones?
Intuition is only a gut feeling, an unsupported conviction that a certain look will attract and hold more attention than an existing look. It's a hunch. And like most hunches, it has as much chance of being right as it does of being wrong.
A/B testing gives you absolute numbers to confirm your ideas, or to get you to reconsider them. You might even be surprised at what you find. For instance, the power of images is well-documented. But a test of a call to action on a Hubspot landing page, reported by Which Test Won, found that removing the image increased submissions by 24 percent. Likewise, a counter-intuitive change of color, from green to yellow on a call to action button, increased lead generation clicks for a company by 14.5 percent.
Every test is unique, because no two audiences are alike.
There also are plenty of tests to consider. They can range from the simple, like the best time of the week for getting people to open your emails, to the appeal of a specific incentive, like a discount on teeth whitening, to new patients. Or they can be complex, like the wording and the placement of a call to action on a landing page, the look of your website, or the effectiveness of a pay-per-click (PPC) advertisement on Facebook.
To know with certainty what grabs the attention of your patients, gets them to engage with your practice and sign up for your services, requires knowing what appeals to them specifically. And that means number crunching, statistically significant testing.
Are you interested in learning more about how A/B testing can help to market your practice? Contact us here.
Posted by Grace M. Frank
Grace M Frank is a freelance writer and editor, and the owner of Frank Communications, www.frank-comm.com.