A/B Testing: Do More With Less – By Zach Kwarta

by Jenae Reid on October 2, 2017

Imagine being able to double your website conversions, sales, or profit without spending another dime on more traffic. A statement like that sounds impossible, or something a snake-oil salesman might promise. But, with A/B testing, it might be closer than you think.FeedFront issue 40 cover

In the most simplistic example, A/B testing splits your website traffic into two different experiences, and measures the differences of things like bounce rate, conversion rate, CTR (Click Through Rate), or time on page.

Believe it or not, simple things like changing your CTAs (Calls To Action), headline copy, and imagery can have huge impacts on your website performance.

Where to start? There are plenty of testing providers that can help get you up and running in only a few minutes. Do a quick search for them on Google – they range from free to pricey. Every solution will be a bit different, depending on your site’s traffic and needs. Most of the tools require only adding a tag on your website to get started, and there are plugins for popular platforms like WordPress to make setup even faster.

After you have an A/B testing platform up and running, the common first question is, “What do I test?” This is where the fun begins.

First, look inside your website analytics, and identify your most common and important funnel path – the one most users go through to generate your websites desired action (clicks on links, sales of a product, lead contacts through your website).

Google Analytics has this functionality built in through a couple of reports. Look for the report called “Behavior Flow” in the Behavior category and set your landing page to the start of your funnel. Alternatively, you can use the “Reverse Goal Path” report, to work your way backwards from the completion of your goal.

I recommend starting with the most leaky part of your funnel, the point where the most users are dropping off. Typically, this is the start of their journey on a key landing page, and if testing proves successful, it can drive significantly more users through your funnel at no added cost.

Now, it’s time to test. You can begin to test the small things referenced earlier in this article to see if they have any immediate impact on your site’s performance. These smaller changes will generally take quite a bit of time to reach testing significance, the measure of confidence that the change is impacting things. Sites with smaller levels of traffic also will fall into this trap.

Don’t be afraid to test major changes. I’ve seen the greatest improvements happen with complete site transformations, new page layouts, or changes to the entire funnel itself.

Most of the time your testing will result in negligible or negative results – don’t despair. Once you find a winning test, be sure to change the traffic allocation to 100%, hard-code the changes when able to do so, and move on to the next test.


Happy Testing!


Zach Kwarta is Marketing Directory by day, nerd by night.  Find him: https://zach.marketing.

This article appeared in issue 40 of FeedFront Magazine, which was published in October 2017. https://issuu.com/affiliatesummit/docs/feedfront-40


Comments on this entry are closed.