CHALLENGE

"I know we're not alone out there, dealing with multiple sites and content versus sales sites," says Noah Parsons, Product Marketing & Sales Director at Palo Alto Software.

Palo Alto got most of its online sales funneled in from two very different sources:

Source #1: Bplans.com content site

Although Palo Alto set up this content site featuring articles and Q&A for entrepreneurs as a marketing ploy, it's definitely not salesy.

"It's trying to generate sales leads obviously, but at the same time be considered a legitimate place to do research and get business information. I don't want to corrupt that too much. We keep the multiple domains, the content side of things, very much separate."

But having multiple domains makes it hard to track how one domain's traffic flows through to the other.

Parsons knew which individual pages on each site were the most popular -- but he didn't know how they affected each other, nor how the sales funnel really worked. Did visitors to particular Bplans content convert better than others?

How could he tweak the content site to make it generate clicks more likely to convert on the ecommerce side?

Source #2: Search engines

Palo Alto's Business Plan Pro software, mass marketed in boxes by retailers across the US, was well enough known that many Web surfers simply typed in the brand name to a search engine and clicked directly through to product info pages on Palo Alto's site.

Again, Parsons knew which pages got the most search traffic, but he wasn't sure how to tweak them so more converted to immediate buyers … and fewer bailed.

CAMPAIGN

Last September, Parsons launched a focused improve-our-site-conversions initiative.

First he used site analytics software (link below) to track the conversion rates of typical visitor paths through the sites. For example, he analyzed:

o People who read general entrepreneurial articles on BPlans
versus people who read business-plan-specific articles. The
content team wanted to continually freshen the content on
Bplans to attract repeat visitors who might be more likely
to convert … but was it better to add yet more plan-focused
topic stories, or to broaden to more topics?

o People who clicked on a PDF sample plan versus an HTML
sample.

Next, Parsons and the site design team applied a combination of Web analytics, a "good old cheap usability group" (gathering 10 typical consumers into the conference room and asking them to surf the site for an hour), and data from past packaging design studies Palo Alto had conducted to improve in-store sales, to figure out how to improve specific page conversions.

They focused efforts on the two most visited pages on the ecommerce site:

o The so-called "decision page" is the first page most people clicking in from the Bplans site see. Its goal is to guide them on to whichever particular product best suits their needs. Since Palo Alto has plenty of different products, this could be confusing.

Should every product be displayed on this initial page, or should it focus almost solely on a single product? What should appear above and below the fold?

o The "product page" is the second-most frequently visited page. It describes one particular product in loving detail - hopefully so compellingly that people decide to buy immediately. The majority of search traffic came straight to this page without seeing anything else. (In fact it gets a lot more traffic than the Company's home page.)

Would search traffic want to see a strong benefits headline, or just the product name at the top of the page? Which were more important -- testimonials written by actual customers or carefully copywritten stories about real customers?

The team tweaked, watched numbers, and tweaked again. (Link below to before and after screenshots.) Then they watched sales over the past quarter.



RESULTS

Sales conversions from the total of monthly unique visitors across both of the sites leapt from 1.01% to 2.39% due to the changes.

10% fewer of people landing on the "decision page" bailed without clicking further into the site. Parsons says, "This high up in the sales process, reducing the exit rate by 10% makes a huge difference. It gets more people down into the sales funnel."

Plus search-generated visitors who landed directly on a particular product page were 50% more likely to convert to buyers on the spot.

Here's some details on factors that made the difference:

- More focused articles won hands down. The editorial team who had been planning to expand the Bplans site with broader topics, now narrowed to add more business-plan specific items.

Crunched for time and resources, they found a clever way to repurpose content by turning popular Q&A content into formal articles with just a few tweaks and a headline change. (Link below to sample.)

- Content site visitors who looked at sample plans were far more likely to convert to buyers than anyone else, however it didn't seem to matter at all if the samples were in PDF or HTML. The key was having a live hotlink to the ecommerce site on each page of the sample.

- "Decision page" visitors were more likely to continue shopping if they were presented with fewer product options -- especially above the fold. Although Palo Alto has lots of great products, the general initial landing page wasn't the best place to display all of them.

Parsons explains, "We were giving them too much information about the differences between our products on that page. They were confused about the Pro versus the Premiere version. People were forced to make decisions about 'Am I a pro or not?' instead of learning about the products as a whole and then making a $100-300 decision later."

So, the vertical design, with various products displayed in more than one column above the fold, was scrapped for a focused horizontal design. Now the content above the fold is focused on one pitch.

- Search engine visitors who clicked straight to an individual product page were confused by the benefit-oriented headline. "They honestly weren't sure what products they were at. It was the weirdest thing."

So, Parsons switched headlines, putting the name of the product at the very top of the page so everyone knew they'd landed at the right place. Then the benefit copy flowed underneath.

- Crafted stories about real-life customers worked far better than testimonials written by customers themselves. Why?

"I think people are getting jaded from marketers going in and writing Amazon reviews," says Parsons. "People don't trust testimonials. They think anyone can make this stuff up. They don't know who Joe from Indiana is."

- Not-so-great photos of the real-life customers won over slick photos. Turns out Web surfers think really professional photos might be clip art, and again, not real. A digital snap taken by a profiled entrepreneur's best friend is much more believable, even if the production values aren't as nice.

While he's thrilled by the revamp results so far, Parsons is quick to note that it's not done. "It's a never-ending cycle.
It would be nice if you could finish a site someday, but…

"Also things don't seem to stick completely. You may see a jump up and it will drop back down again a little. It won't go down the original level though -- it's two steps forward and a small step back. We're really happy, but we're going to keep working on it tooth and nail."

Useful Links related to this article:

Before-and-after screenshots of tweaked pages, plus a Q&A turned into an article:
http://www.marketingsherpa.com/pa/ad.html

BPlans - the content site Palo Alto developed for prospecting
http://www.bplans.com

Palo Alto
http://www.paloalto.com

WebSideStory - the site analytics software Palo Alto used to figure out which tweaks would raise conversions
http://www.websidestory.com