A/B Testing: The Top Five Things to Test

Article written by:
  • Inbound Marketing
  • Updated: May 19, 2015

Website usability and conversion testing is one of the easiest ways to improve the profitability of a site yet it’s often overlooked as many of us get too hung-up on driving the largest amount of traffic possible to our sites. Yet what’s the use of traffic if it doesn’t convert, after all for most us achieving conversions (be that sales, banner clicks, newsletter sign-ups, enquiry forms or affiliate product sales) is what pays the bills. A/B testing is one of the simplest ways to increase your conversion rate, this article looks at 5 of the top elements to test, if you need a generic overview of how this technique works there’s more here in this guide from A/B testing software provider Maxymiser and you can make use of Google website optimizer to A/B test for free.

Marketers know that in an A/B test, just one element is changed for the duration of the test to achieve clean results against the control piece. If you change too many elements are once, you’re either straying into multivariate testing or you’re muddying the waters so much you won’t be able to get a clean result.  Yet if you can only change one element, which element do you choose to get to the most profitable results quickly?

A/B Testing Guides
(Courtesy of Flickr User: mil8)

The Top Five Items to Change and Assess in an A/B Test

1. Creative

The creative element refers to the graphic design or layout of the piece. It may be as simple as changing the color (green to orange) or the font (Verdana to Times New Roman.) You can try changing one aspect of the overall appearance of the piece or the entire design.

2. Headlines

Headlines are the most important aspect of the copy, and testing several headlines often yields a quick bump in response and conversion rates. Try testing a question (How Happy Are You With Your Bank?), a benefit statement (Better Banking for Less – New Lower Monthly Fees) or a provocative statement (Most Banks Are Blah).

3. Call to Action

The call to action is the statement near the response mechanism that directly asks for the lead, sale or action step you want customers to take. Try tweaking the wording in one test, the placement in another. Sometimes the call to action is obscured by the graphic design or the wording is weak. A simple change has the potential to boost response rates with minimal changes to the overall piece. However, you won’t know until you test it!

4. Timing

Timing is often an overlooked aspect of  A/B testing, but it can have a surprising impact on response rates. If you typically post to your blog in the morning, test the results of morning versus evening posts (there are automated features that let you write posts in advance and schedule the timing). If you typically send email marketing out on the first of the month, try the first Monday, or perhaps Tuesday. Maybe Wednesday is the best day for your customers because none of your competitors are sending out messages. The point is that timing is one aspect of your marketing efforts that’s typically inexpensive to change and may boost response rates. It’s worth testing.

5. Offer

What are you offering your customers? Are you trying to get them to sign up for a magazine, buy a book, order a free information kit or take some other action? Change the wording of the offer. If you give away a premium or a free shipping incentive, test one against the other. Does the premium generate a higher response rate or the free shipping incentive? Modest changes in how the offer description is worded may clarify the offer further or entice customers with more appealing language.

No matter which element you test, be sure to roll out your A/B test at the same time.  Don’t send out your control first, then your test. Even the difference of a week or two between the pieces can alter the results so much as to make the test untenable.  Change one variable per test for your A/B testing efforts, and be sure to have a clear-cut method of tracking the responses.  Soon you should have enough data to tell which among your test piece is the clear winner.

Article by WHSR Guest

This article was written by a guest contributor. The author's views below are entirely his or her own and may not reflect the views of WHSR.

Get connected: