Exclusive Wrap-Up Report: Conversion Conference 2010

by Natalie Tomasso, Senior Reporter, WhichTestWon.com

Office Politics, Traffic, and SEO Top Concerns Among Conversion Conference Attendees

Last week ~200 marketers attended the first ever Conversion Conference, held in San Jose. Given the fact that the conference was dedicated to improving conversions, we were surprised when keynote Tim Ash asked “who’s testing?” and only half of the hands went up. It’s a heck of a lot better than industry average (about 26%) but, still, we hoped more people would be actively testing.

“Are you testing? And, if not, why not?” we asked attendees on the show floor. Here’s what they told us.

Attendees’ Top 5 Testing Challenges

#1. Office Politics.

We heard lots of anger and frustration on this point. “We want to test, but it’s getting the hierarchy on board,” one ecommerce marketer told our reporter. “The department heads think they know all the answers. And, you have to go through them to implement anything.”

“It’s hard to convince IT to support testing,” said a marketer from an online lending site. “And they have higher pull in the company (than marketing).”

“My CEO does what he wants,” a travel marketer told our reporter. “And, it’s hard to introduce something new when everybody’s plates are full.”

Even big companies have trouble. “Even though we’re very numbers driven, people (in the company) still don’t understand analytics,” said a marketer from a huge consumer electronics retailer that conducts hundreds of tests per year.

(By the way, if office politics is a problem for you, you might be interested in this WhichTestWon report. Here’s the link: Office Politics & Budgeting Battles on Testing)

#2. Traffic.

Marketers continue to have issues generating enough traffic to the pages they’re testing. Dale Stokdyk, Internet Marketing Manager, Champlain College, told our reporter, “Our challenge is generating enough page traffic to get conclusive results.”

For pages with low traffic, we think reducing the number of test variations can help in reaching conclusiveness. If you’re trying to test a template page though, where each page may have few visitors but a lot of pages on your site use that same template, you may want to get outside help in designing and running a conclusive test.

#3. SEO.

Despite the fact that, in the past, Google has announced testing won’t affect natural search rankings and presenters at the show said it’s pretty much a non-issue, many show attendees expressed concern. In fact, someone at nearly every Q&A asked how not to lose SEO ranking due to testing. We suspect SEO consultants seeking clients could use ‘I know how to make testing safe’ as a great marketing hook. In the meantime, the good news is there are things you can do to prevent testing from damaging your SEO rankings. Here’s a link to a WhichTestWon presentation all about the topic : Testing SEO Landing Pages: How to Avoid the Risks.

#4. Technology.

There are plenty of testing technologies out there, but not all play well with sites that have complex conversion patterns and structures. And people with complex sites are often the ones most eager to be testing. Technology complaints at the conference ranged from difficulty integrating free testing tools with ecommerce platforms to even more complex problems.

“One problem is getting the technology to pre-populate test registration forms with data already entered on previous landing pages,” a marketer from an online dating company told our reporter. “You can’t build application logic into it.”

One ecommerce marketer said he’d like to test multiple shipping offers but can’t find a testing technology that will recalculate shipping costs by offer. Sites that segment everything want technology to slice results data so they can see winners from different types of traffic.  “We have to go into discover and slice it ourselves,” said the online dating marketer.

#5. What to Test.

Everyone wanted to know which pages and page elements were most worth testing, and what might not be worth the effort. After all, no one has unlimited testing resources.

Among the laundry list of things people wanted to test were short copy vs. long copy, simple page design vs. complex page design, and “buy now” call-to-action buttons. However, according to the case studies presented at the conference, here’s what’s really worth testing if you can only run a limited number of tests….

(Also, if you’re having trouble deciding what to test, check out WhichTestWon’s report: What’s Worth Testing & What’s Not)

Top 3 Tests Worth Running — Case Studies from the Conference

A. Test trust icons and authority icons.

Security icons, payment icons, and privacy icons all play a huge role in reducing anxiety and building trust. We’ve seen a single security icon (i.e. a lock with a checkmark on it) increase actual purchases by as much as 20% on a product checkout page. Similar results were shared at the conference.

Adding authority icons such as “As seen on” media logos like CNN, MSNBC, and Good Morning America contributed to a 40% lift in registrations for RealAge, an age assessment and health site.

This is a pretty basic tactic, but surprisingly few marketers are accurately using trust and authority icons to get visitors to purchase. It may not be as exciting as a total page revamp or short copy vs long copy … but it works.

B. Reduce page load time.

Is speed a variable worth testing? It is for Mozilla, which conducted an A/B split test on load times for its Firefox homepage. They reduced page load time by combining files and rearranging the source of the two test pages. The results: Reducing page load time by 2.2 seconds increased downloads 15.4%. And with about 16 million unique visitors per month, that equates to millions of new browser downloads.

C. Test shopping carts and lead generation forms.

According to case studies and anecdotes we heard at the show, copy and creative are less important than testing your shopping cart and lead gen forms. The check-out or fill-out process is where you’ll get the biggest bang for your testing buck.

For example: athenahealth, an online physician billing provider, increased lead gen form fills 84% and reduced bounce rate 48% on a PPC landing page by switching from a single page to a multi-page form. The multipage form asked visitors to select from four health care roles on the first page, to select a practice size on the second page, and fill out an 8-field form on the last page in order to get a free PDF.

Here’s a link to the Conversion Conference’s site: http://www.conversionconference.com

Got testing anecdotes of your own to share? Join the discussion by posting a comment below!