The Entrepreneur Network

Online merchants have until Friday to participate in a contest on the site. We sat down with Anne Holland to learn more about the site, the contest, and the idea behind the concept of A/B testing which concentrates on helping sellers boost conversions on their websites.

Anne Holland founded MarketingSherpa, a privately held media company, in January 2000, which she sold to MarketingExperiments.com in 2006. Holland is now President of Anne Holland Ventures Inc., publisher of WhichTestWon.com, an inspirational site for online marketers on A/B and multivariate testing, featuring a weekly real-life test to vote on.

What is WhichTestWon.com and who are your readers?

Anne Holland: We’re the only weekly 100% dedicated to conversion optimization and A/B testing. Our editors have researched and created an exclusive online library online of more than 300 A/B tests including screenshots and results data from many ecommerce sites, as well as lead generation sites.

Each week we publish a new A/B test (with results data) that you can check out for free on our homepage. Our readers are marketers around the world who want more of their site traffic to convert. They use our Case Study examples for ideas of what they can change or test on their own sites – including their cart, product pages, buttons, images, etc. – to get higher conversions from the same traffic.

What is A/B testing and why is it important?

Anne Holland: A/B testing is the only safe and scientific way to improve your own site or ecommerce offering conversions. You can do it on a web page, an email, an opt-in form, a blog, an online ad, even on Facebook ads. You simply split the regular traffic to your page so half of the traffic sees one version of the page and the other half sees a different version. Then you sit back and watch to see which version makes more sales… and ultimately implement the winning version! Lots of very cheap (and even free) software is available to run A/B tests these days. Most allows you to just insert one snippet of code onto your page and the two versions start running. Then the software does the math for you, telling you which page is the winner.

According to our Case Studies, most ecommerce sites can get at least 22% more sales – from the same traffic – if they run a few A/B tests. Lead generation sites can improve results by an average of 40%. That’s why companies like Amazon and Dell run A/B tests continually on their own pages to see if they can tweak anything from their images to their copy to improve sales. The good news is, even small companies or mom-and-pops can run tests these days because the software is so cheap and easy to use. We have a free guide to the various tools you can use.

What tools do a small online merchant need to run A/B tests on their ecommerce website and how do they measure success?

Anne Holland: The tools measure success for you. When you run an A/B test, look for a result that is at least 90% conclusive. That means the winning page is statistically safe to use as your regular page. Google Content Experiments is a free tool, and there are many inexpensive ones available as well.

What kinds of things should merchants test?

Anne Holland: Some of the most successful tests are the big obvious stuff – making your product image larger, changing the wording of your headline, changing your “add to cart” area or button. It can also help to simplify your page design, remove extraneous navigation, add a security seal in your cart, etc.

What are some of the biggest mistakes sites do wrong when it comes to converting visitors to buyers?

Anne Holland: Well, not testing is the biggest! Your “gut” about what the best page design is, or how your cart should look, or what your page headline should say, is very often wrong… because you are not your target audience so you can’t know what will make them buy until you test things. Also, sometimes best practices in design are wrong for your audience. That’s why even companies like Yahoo test their pages and their email templates. They know they can’t get it right without testing.

Smaller companies often have too much clutter on their pages, and they generally do not have big, obvious-enough buttons. I often find smaller companies may use several headlines and sub-headlines because they aren’t sure what their message should be. Instead, they should find the perfect wording through testing and stick with one big perfect headline and maybe just one perfect sub-head.

Tell us about the contest you’re running and how merchants can participate.

Anne Holland: If you’ve run an A/B test in the last year and it was at least 90% conclusive, you can enter WhichTestWon’s 4th Annual Testing Awards for free. Entry forms are on the WhichTestWon.com website – but the deadline is this Friday Jan 25th, so hurry! You can also sign up at that page to see who won when we make the announcements. It’s our most popular editorial of the year – tens of thousands of marketers go view it.

Where and when will you be publishing the results?

Anne Holland: We’ll do a live webcast for the awards ceremony in February (you can sign up for free). Then we post winners to our Hall of Fame. Past years’ winners are listed at WhichTestWon.com.