August 24, 2006
Does your home page already have pretty good, clean design? CareerBuilder.com's home page for employers did. Then, a newly promoted marketer decided to change it.
Discover his A/B test results (potentially worth millions in business account revenues). Includes the answer to the big design debate -- text-heavy links vs short copy with
Have you ever dreamed of testing radical changes to your company's home page, if only you had the power to do so quickly and easily?
After three years of leading one of CareerBuilder.com's tech teams, Michael DeHaven grabbed the chance for a new career by switching departments to head up ecommerce marketing for the company's online sales to employers.
"The home page I inherited when I took this job over was not a bad page at all," he explains. In fact, it was very cleanly and professionally designed. (Link to sample below.) The cheerfully colored page featured a photo of a smiling woman next to big fat click buttons for each of the four main navigation options:
- Post one job now for $389
- Buy a 4-pack for $1,299
- Search resumes for $500
- Sign-in access for current account holders
"It's pretty clear what's going to happen when you click on one of those links," says DeHaven. "The issue was after you clicked. We had these huge drop-off amounts."
How can you redesign an already-ultra clear site to improve conversions?
First, DeHaven considered the overall strategy behind the navigation. "The buttons took you into the purchase process with very little information. After all, everyone knows what a job posting is, right? My first question was, is that correct?"
So, he decided to test adding a "Learn More" hotlink next to each "Buy" button to see if anyone clicked on the former and if they then converted.
DeHaven also reviewed the layout and design of the page with a tester's eye. He decided to test three significant changes:
#1. Text-format with textual hotlinks instead of colorful graphics with click buttons.
#2. Main options listed in a vertical column instead of a horizontal row.
#3. Moving the user sign-in form to a small corner in the upper right of the screen, rather than allowing it more valuable real estate.
DeHaven ran the test for a week, splitting traffic to discover which page converted more visitors into sales. He also examined back-end data to see if sales reps were having better luck converting leads that had been to the test version of the site.
Then, DeHaven used results to create one more test panel. This time he tweaked all the hotlink wording on the page to see if longer, wordier hotlinks that search engines love would be better for human beings as well. (Link to sample below.)
The design team didn't think much of this test. "Everybody who saw it internally said version C was way too crowded and there wasn't enough white space. We worried people would see it and feel overwhelmed with links."
Turns out everyone was wrong. The heavy-text version that got thumbs down internally won more customer accounts than the cleaner, more graphical design.
It seems that business executives prefer to look at fairly plain textual content online rather than cheerful graphical interfaces. Plus, they prefer vertical to horizontal groupings of options and longer, wordier textual click links.
"After I thought about it awhile, it made total sense," explains DeHaven. "Users are trained to allow their eye to scan down something that looks like search results. And what we have here now is something that looks an awful lot like search results. It's a good quick scan. Graphics and images are not what the eye's trained for online.
"It was a HUGE finding. Really exciting."
In fact, the final winning version worked so well that there was a 13% decrease in the time the average visitor spent on the page before they clicked on a navigation choice. This indicates visitors were less confused and more likely to be happy with the clickpath they entered … and hence more likely to convert at the end of it.
Orders per unique visitor rose 7% overall, including immediate sales and bill-mes. That's tens of thousands of more potential revenue per month.
A few more interesting findings:
- When presented with the "buy now" vs "learn more" option, about 40% of clicks were on "learn more." These didn't always convert as well as the 'buy now' clicks, but 'learn more' registrations did produce valuable sales leads for the sales force.
- You should hotlink your headlines and product names as well as your official click links. Although headlines were not clicked all that often, they were clicked enough to be worth hotlinking.
- Moving the sign-in form for existing accounts to a less noticeable part of the page had absolutely no impact. Lesson learned, don't waste valuable marketing real estate on account holder logins.
Useful links related to this article:
Creative samples of the three home page tests:
Omniture -- the Web analytics firm CareerBuilder.com uses:
CareerBuilder.'s current home page for employers