May 24, 2007
If your conversion rates are in the doldrums and you know you need to spice up your offer, a multivariate test might help get the boost you’ve been hoping for.
Here’s a real-life Case Study that shows how one subscription site tested headlines, banner ads, pricing, bonus offers and six other variables. Best of all, a higher price came out on top. Includes before-and-after images and results.
“I’m facing one problem just like any other ecommerce practitioner -- the conversion rate,” says Jodjie Azurin, General Manager, BusinessSummaries.com. The subscription-based website, which offers busy executives summaries of popular business books, had seen its subscription conversion rate decline for several years, falling to 0.019%.
Azurin and his team knew they needed to focus on their subscription offer page and tried several copy rewrites, but any lift was only temporary, with conversions slipping back into the basement. They needed a better way to analyze the components of their subscription offer page, then find the right levers to boost conversion rates and keep them high.
Azurin had read about multivariate testing, so after his team’s in-house efforts failed to deliver a sustained conversion boost, he turned to a vendor (see hotlink below) to set up a rigorous multivariate test of their subscription offer page. Here’s what they did:
-> Step #1. Analyze visitor behavior patterns and customer service queries
Before setting up the test, they worked to understand exactly what users were doing at the BusinessSummaries website, looking for patterns that might indicate problem areas or other challenges that prevented users from signing up. They analyzed the following factors:
- Emails sent to the BusinessSummaries customer service team featuring questions about the website, subscription rates and other terms and conditions.
- Traffic statistics, such as what pages visitors viewed and how long they stayed at the site. By looking for pages where a high percentage of users abandoned the site, they could focus on elements to tweak during the multivariate tests.
- An exit pop-up survey asking visitors for their primary reasons for not ordering or particular objections or questions that users had about the service.
-> Step #2. Research competitive sites
Azurin has seen an increase in competition since the company's founding in 2001, so the team analyzed the subscription offers of those competitors, looking for differences in how those firms organized their sites and subscription offers, as well as pricing levels.
-> Step #3. Identify test elements
From that market research and site analysis, they focused on 10 variables to test.
#1. Headline. Site statistics had found that a large number of visitors abandoned the subscription offer soon after clicking, leading the team to wonder whether users were not getting a compelling reason to continue with the subscription process. This led them to develop two new headlines: “Who Else Wants To Phenomenally Boost Professional Success And Leave Competitors Far, Far Behind?” aimed at Entrepreneurs and Employed Professionals, and a headline that listed the titles of a handful of clients from major brands, such as Coca-Cola, and said, “Discover The Closely Guarded Secrets Used By These Leading Executives To Super-Charge Their Profession, And Stay Ahead Of Others In Today's Fast-Changing Business World.”
#2. Banner at the top of the page. This was another test to capture those visitors who were clicking away quickly. The existing BusinessSummaries page was text-heavy, with no graphic banner element at the top of the page. So, the team designed two new banners to load with the page -- a static image and one with Flash animation.
#3. Subscription offer box. This section outlined the names of books available in the database, making them clickable links that allowed visitors to sample an excerpt from that book’s summary.
#4. Entry pop-up offer. The existing site used a pop-up offering users a free, two-day trial of the service and four free summaries. The team tested two other pop-up offers: a 15% discount if the visitor subscribed within two hours and 15 free summaries with no free trial period.
#5. Audio message from the site’s owner. Some users had requested an audio component of the subscription page, so the team tested two audio messages: one allowed users to click to play the message; the second started the audio message automatically after the visitor had been on the page for a few seconds.
#6. User testimonials. The existing site featured a few user testimonials on the subscription page, so the team created a second version that increased the number of testimonials.
#7. A “Book of the Month” feature. This was a new element the team designed to highlight a particular book in the company’s database.
#8. Images of book covers. The existing version had no graphics of books covers, so the team created a version that provided some images to see if the visual representation of available titles would affect conversions.
#9. Bonus offers. The existing site offered a few free gifts when users bought a subscription. The team created a second version that increased the number of gifts.
#10. Price. The default price for a one-year BusinessSummaries subscription was $69.95. Two variables were offered that increased the price by $10 or decreased the price by $20 to see which price point had the best conversion rate.
-> Step #4. Implement test pages with live Web traffic
The testing algorithm served up 9,000 combinations of those 10 variables, along with the existing subscription offer as the control page, to live Web viewers. Given the site’s low conversion rate, they needed to serve test pages to 110,000 visitors over a few weeks to get a reliable winning version.
The winning version of the subscription landing page delivered a 697% increase in revenue per visitor over the default page and surpassed everyone’s expectations. Needless to say, Azurin is "very, very happy with the results."
Besides achieving their goal to boost subscription conversions, the test also gave Azurin’s team valuable insight into the most important elements of their subscription offer page. Most significantly, they learned that increasing subscription rates $10 actually increased conversions. “It placed some sort of a premium in terms of perception of our offer. We priced our offering too low to begin with. Maybe at a later date, we can test an even higher price.”
A strong headline helped convince visitors to buy a subscription. The headline that tested best was the one for entrepreneurs and employed professionals, which promised to “Phenomenally Boost Professional Success.” Surprisingly, though, adding a banner to the top of the page decreased conversion rates, with the team speculating that it distracted visitors from the copy below.
- Adding an audio message actually hurt conversion rates, despite the fact that some people had asked for it.
- Increasing the number of testimonials did not increase conversion rates. The team stuck with their default use of testimonials but are considering reducing the number of testimonials in a future test.
- Elements that provided more information about the summaries available to subscribers -- the offer box with descriptions of titles in the database and samples from the summaries and the Book of the Month feature -- increased conversions.
- Adding images of the book covers did not increase conversions, which is perhaps a reflection of the nature of their audience, who may react more to text than to graphics.
- The pop-up promotion offering a 15% discount to subscribers who signed up within two hours delivered more conversions. By contrast, increasing the number of free premiums for new subscribers did not lead to more conversions, so the team stuck with the default bonus offer.
Useful links related to this article
Creative samples from BusinessSummaries.com’s multivariate test:
Conversion Multiplier - BusinessSummaries.com’s multivariate testing vendor: