by
David Kirkpatrick, Senior Reporter
CHALLENGE
Website optimization can be a powerful driver of conversion, and even testing a single element can prove valuable. For an e-commerce website, removing anxiety from the visitor's path, which might impede a sale, should be high on the list of website design goals.
In this case study, we look at that very scenario.
ModernCoinMart, a division of John Maben Rare Coins, Inc., outsources its Web hosting to a third party. While the hosting company scanned the e-commerce site for security issues such as viruses or vulnerabilities, ModernCoinMart did not have visibility into those reports.
"We also questioned if the protection provided by the Web hosting company was sufficient," said Wayne Leiser, Chief Technology Officer, John Maben Rare Coins, Inc.
He added that ModernCoinMart was "always looking at ways to boost traffic" and "increase the number of sales conversions."
"We knew that some visitors to the site were browsing, and considering buying or selling coins, but abandoned their purchases before finalizing transactions," Leiser stated. "We wanted a way to reassure customers that their credit card and personal information was secure."
The result of this marketing pain point was to enlist the services of a separate website security vendor that could provide ModernCoinMart reporting on the security of its website, as well as a trustmark that might reassure customers of that security.
Read on to learn how testing this simple website element led to a 14% lift in website sales conversions.
CAMPAIGN
The website security vendor offered ModernCoinMart a 30-day trial of its product.
Leiser said this trial period gave the company an opportunity to understand the vendor’s website scanning and reporting, as well as offered an opportunity to perform an A/B split test on how ModernCoinMart’s customers responded to the vendor’s trustmark on the website.
Step #1. Analyze website for optimal trustmark placement
The ModernCoinMart website was analyzed for optimal placement of the tested trustmark.
In this case, determining "optimal placement" did not involve any website testing.
The team’s goal was to simply find a prominent location on the website where visitors can see the trustmark above the fold and where they would not be required to scroll down before seeing the trustmark.
The ModernCoinMart website featured a banner with the company logo on the extreme left, and navigation elements sat on the right. Below those navigation tabs were other elements, including a small contact information box and a grouping of social media icons.
The
trustmark was placed in a blank area to the left of the social media icons, giving the trustmark a prominent position in a largely empty banner at the top of the ModernCoinMart homepage.
Step #2. Use special website code to serve the trustmark
The main "under-the-hood" element of creating the website test was a special "survey code" used to serve the trustmark to website visitors on the ModernCoinMart website.
This code is different from a website code that would be used to serve the trustmark on a permanent basis.
Step #3. Split the test subjects with a website cookie
The second "under-the-hood" element was a persistent cookie that was involved in actually executing the test.
Every website visitor was given this cookie, and the cookie alternated with each visitor as a "yes" cookie or a "no" cookie.
For example, visitor number one would receive a "yes" cookie. Visitor number two would receive a "no" cookie, with visitor three receiving a "yes" cookie, and so on through the duration of the test.
This cookie allowed the test to evenly split website traffic between the original site and the ModernCoinMart website featuring the trustmark.
Step #4. Execute the test
After determining the trustmark placement, creating the special survey code, and setting up the cookie protocol, actually executing the test made up the last stage.
The test ran in the fall of 2011 for 30 days, splitting website traffic between the original ModernCoinMart website, the control in this case, and the website with the trustmark, the treatment.
The only difference between the control and treatment was the trustmark, with it placed in an area of the treatment that would have been blank space for visitors to the control site.
ModernCoinMart visitors who received the "yes" cookie were directed to the treatment, and "no" cookie visitors viewed the control website.
RESULTS
Over the 30-day website test, Leiser said, "Half of our 74,890 visitors to the site saw the trustmark on the site, and half did not."
With this visitor count, the statistical confidence level for the 30-day test reached 90%.
At the end of the thirty days, the treatment website featuring the trustmark beat the control with a 14% increase in sales conversions.
"The A/B testing showed that the trustmark was beneficial," said Leiser.
He added, "People are familiar with (the security vendor’s) logo. Even if there wasn’t an increase in sales, the comfort of having the website scanned daily for any vulnerabilities by a professional company like (the security vendor) is well worth knowing that we are taking yet another step for our customers’ safety and security."
For more consumer marketing case studies and how-to articles, subscribe to the bi-monthly MarketingSherpa Consumer Marketing newsletter.Creative Sample
Security vendor trustmark on ModernCoinMart’s homepageSources
ModernCoinMartJohn Maben Rare Coins, Inc.McAfee – ModernCoinMart’s website security vendor
Related Resources
Website Optimization: Simple A/B test adds 'thousands and thousands of dollars' to Cars.com’s bottom lineA/B Testing: 4 tests from a crowdfunding site with double-digit resultsA/B Testing: Working with a very small sample size is difficult, but not impossibleA/B Testing: How a landing page test yielded a 6% increase in leadsWebsite Optimization: 1,915 marketers rank the top optimization goals
