MindValley Labs Internet Marketing Blog

Lets face it. If you are a small busi­ness owner, you are not the worlds’ best web designer or copy­writer and you can­not afford to hire them. For­tu­nately, now that you have Click­Muse you don’t need to.

I can­not tell you how many times I have stared at my own web­sites and won­dered whether I should go with ver­sion A or B or how many days I have spent try­ing to come up with the “per­fect land­ing page.” Then I dis­cov­ered a bet­ter, faster, sim­pler way. Instead of ago­niz­ing over what head­line to choose or whether to go with ver­sion A or B of a page, why not just quickly split test them?

What is split testing?

Split test­ing is an extremely pow­er­ful online mar­ket­ing method that allows you to test two ver­sions of a web page at the same time. This is why it is also fre­quently called A/B test­ing. When­ever you are split test­ing a page, you have two versions:

Ver­sion A: is the con­trol group. This is your exist­ing web page

Ver­sion B: is the new test

Your goal in split test­ing is to always try to come up with a new test to beat your con­trol group. As soon as you have cre­ated a new ver­sion to test, Click­Muse will start to serve the two ver­sions of your page in real-time and show you which page is per­form­ing bet­ter accord­ing to your defined goals (such as sign-ups and sales).

The art of split testing

To dra­mat­i­cally boost the per­for­mance of your site with split test­ing, just keep two prin­ci­ples in mind.

1) Test everything

2) Never stop testing

When you are first get­ting started with split test­ing, there are a cou­ple of obvi­ous tar­gets that you can imme­di­ately get started with. The obvi­ous tar­gets are site ele­ments that are most likely going to have the biggest impact on your site’s con­ver­sion rate. We have high­lighted the top 6 tar­gets for split test­ing in the image below.

Top split test­ing elements:

1) The top header: The top header of a web­site is the first ele­ment that loads on the page and the first thing that your vis­i­tors will see. Our tests have shown that sim­ply vary­ing the color or infor­ma­tion that is pre­sented as part of the top header can have a huge change in the response rate.

2) Main title: As we already cov­ered in the first les­son, the main head­line is the sin­gle most impor­tant part of your site and you should always try to come up with a bet­ter head­line than the one you are cur­rently using.

3) Sub title: Often ignored is the sub­ti­tle. How­ever, we have had tests weree vary­ing the sub­ti­tle alone increased our sign-up rate by over 135%!

4) Body text: It is copy not images that sell, so test­ing dif­fer­ent copy ideas is absolutely crit­i­cal to max­i­mize the poten­tial of your web­site. Pay spe­cial atten­tion to the text that shows up right after your head­lines since this is the text that shows up above the fold. We have learned that often list­ing the main ben­e­fits as bul­let points can lead to supe­rior results instead of lead­ing in with an open­ing paragraph.

5) The hero shot: Every site will have at least 1–2 images on your main land­ing page. Often, as in the exam­ple above, the hero shot is used to call atten­tion to the most wanted response (such as the sign-up form). Hero shots can have a major impact on your site’s per­for­mance and are one site ele­ment def­i­nitely worth testing.

6) The sign-up form: If your most wanted response is to cap­ture leads (i.e. email addresses) then you have to test dif­fer­ent sign-up forms since that will have an imme­di­ate impact on your sign-up rate. Split test­ing dif­fer­ent forms, col­ors, text, and posi­tion­ing of your form will lead to sig­nif­i­cant improve­ments over time.

There are two ways to run split tests:

1) Split test­ing one ele­ment at a time: The eas­i­est way to split test is to pick one ele­ment such as the head­line or the sign-up form and to run and A/B test to test two dif­fer­ent ver­sions. As soon as you have col­lected enough data, you will know which ver­sions is per­form­ing bet­ter and then you can either run another split test of the same ele­ment or move on to test­ing some­thing else. This test­ing approach is very straight for­ward and effective.

2) Split test­ing mul­ti­ple ele­ments: Instead of test­ing one ele­ment at a time, you might also want to test mul­ti­ple vari­ables at the same time or test a com­pletely new design of your site. The main advan­tage of this approach is that it allows you to rapidly test major changes to a web­site and to test mul­ti­ple vari­ables at the same time. The dis­ad­van­tage is that at the end of the day you will know whether ver­sion A is bet­ter than ver­sion B but you will not be able to pin-point it to one exact data point.

At the end of the day, your goal is to con­tin­u­ously keep test­ing and to con­tin­u­ously run two ver­sions so that you keep mak­ing your site bet­ter and bet­ter. I hope that the above has given you some good ideas of what main ele­ments to start test­ing on your site today.

Last 5 posts by

• Do you think SEO in other lan­guages is eas­ier? — Sep­tem­ber 18th, 2008
• Google Slap — How to recover from a Google Slap? — Sep­tem­ber 8th, 2008
• What Does Your Order But­ton Say? — August 29th, 2008
• Will Knol Rule The Search Engines? First Analy­sis Indicates…It Just Might — July 23rd, 2008
• Are small busi­nesses drop­ping out of AdWords? — July 23rd, 2008

About the Author

Prior to MindValley Media, Mike was the Head of New Ventures Strategy at eBay where he conceived of the strategies that led to the investment in Craigslist, the launch of Kijiji.com and the acquisition of Skype. Mike has an MBA from Stanford and previously worked for the Boston Consulting Group. He is also a certified Google AdWords Professional.

Check out other posts by Mike Reining

If you want to see what's in the private lab...

Ask yourself... which part of your business would you most like to improve?

I can send you 7 tactics related to your specific needs, if you like.

You don't have to buy anything, just take it as a backstage pass into our private course.

Will it give you the solution you've been waiting for?

There's only one way to find out.