Testing Is At The Heart Of Marketing Success

Marketing for any business is an ongoing effort that requires consistent refinement and improvement to see positive results in the long-term. It is not enough to simply create a quality marketing program and leave it alone to do its job. Rather, the marketing plan for any business needs to be monitored on a daily basis […]

Chat with MarTechBot

Marketing for any business is an ongoing effort that requires consistent refinement and improvement to see positive results in the long-term.

It is not enough to simply create a quality marketing program and leave it alone to do its job. Rather, the marketing plan for any business needs to be monitored on a daily basis in order to make sure maximum returns are achieved.

Testing, Testing

Thanks to A/B testing, or split testing, it has never been easier to refine marketing programs. Businesses that aren’t using testing as a major part of their marketing plan are missing out on a big opportunity to improve results and grab a larger portion of their desired market.

The idea behind A/B testing is that it can provide real-world answers to questions that may be argued in a marketing meeting behind closed doors. Instead of having to pick between two different designs for new marketing materials, or between two different versions of copy for a website, both versions can be used and tested to find which performs better.

By taking the guessing out of the equation, businesses can be more sure that they are taking their marketing strategy in the appropriate direction.

How A/B Testing Works

So what does this look like in the real world? The following example will help to demonstrate how A/B testing is implemented online by the average business.

  • When developing a sales page for a new product, two versions are created including different copy, color scheme, call to action, graphics, etc.
  • Traffic is sent equally to each of the two versions of the page
  • Using tracking software, the results of the traffic are monitored to see which version of the page is performing better
  • Once a ‘winner’ is determined, that page can be used to serve all traffic going forward

While that is a basic outline of how A/B testing works, the practice of it can be far more complex. Rather than just having two versions of the same page, a marketing department could work up many different variations with tweaks to each small section of the page.

This kind of multi-variate testing can include as many different versions as a business would like to create. The more details that are refined and the more trials that are conducted, the more efficient the marketing strategy can be.

Grammarly Improves Conversions With An A/B Test

At Grammarly, we are committed to split testing and the impressive results that it can offer. There are frequently three or four marketing channels or initiatives at any one time that are being tested and refined. By understanding that the failure of some marketing efforts is an opportunity to build more successful ones, Grammarly is able to improve conversion rates by between 10 and 20 percent.

For example, we recently hypothesized that reverting to a simpler payment form would make our payment page more credible to potential customers who are sensitive to providing personal identifying information. Our default payment page was a hand-illustrated credit card form that jived with our site’s design team, but had a somewhat unconventional layout.

Grammarly, default, landing page, A/B test

(Click to enlarge.)

 We made slight changes to the layout of the page so that its appearance mirrored more typical payment pages. The new layout included a standard, vertical payment form without any illustrations or extraneous design elements.

Grammarly, landing page, A/B test

(Click to enlarge.)

Believe it or not, the new version of the form improved our conversion rate by three percent, at a 98% confidence level. Given how much traffic Grammarly generates to our payment page, that’s a meaningful gain.

Avoiding Over-Confidence

A common mistake is for a marketer or business owner to think that they know what their customers want, and to not bother with A/B testing. In fact, if you are doing it right, you will assuredly be surprised about what works best from a marketing perspective. Luckily, technology makes it incredibly cheap and easy to run these kinds of tests.

Some other things to consider in regard to A/B testing:

  • Intuition Can Often Be Wrong. No matter how sure one is that a particular marketing material is a winner, it is always worth the time and effort to test it with real customers.
  • As With Any Scientific Test, It’s Always wise To Start With A Hypothesis. Knowing what is expected is a good place to start from when evaluating results.
  • There Is No Reason To Ever Stop Testing. As long as the ROI is in line with business objectives, running split tests on an ongoing basis is the best way to seek lasting growth and refinement.

A/B testing, whether simple or complex, should be an integral part of any business. From landing pages and sales funnels to banner ads and more, split testing is a powerful way to have potential customers provide clear answers to difficult marketing questions.


Opinions expressed in this article are those of the guest author and not necessarily MarTech. Staff authors are listed here.


About the author

Yuriy Timen
Contributor
Yuriy Timen is the head of online marketing at Grammarly, an automated proofreader.

Get the must-read newsletter for marketers.