“Everyone is talking about collaboration right now,” said Maria Colacurcio, VP-marketing at Smartsheet.com. “It’s a very crowded market.”
Smartsheet.com’s search marketing initiatives were already fairly sophisticated. It had gone beyond the use of generic search terms to specific keywords.
Smartsheet.com also created more than 40 landing pages for the Web site and became very specific about the markets it wanted to penetrate.
“We went from a very generic search marketing approach to a very specific strategy,” she said. “We incorporated very specific vertical terms that made sense to specific individuals.” Eight segments were targeted: marketing, human resources, operations, nonprofit, real estate, architecture, product management and client services.
The new strategy gave Smartsheet.com a 4% to 6% conversion rate, up from 2%. While Colacurcio called that approach successful, she had no idea which site elements drove the conversions.
Colacurcio next enlisted the help of Widemile, a Web testing and optimization technology company.
“We wanted to test the landing pages, because at that point we’d invested in the strategy,” she said. “We wanted to make sure we were getting the most bang for our buck and the best conversion rates.”
Together, they chose five out of the 40 landing pages to test. But before launching formalized multivariate testing, Widemile suggested ideas for “low hanging fruit”-types of tweaks to be made to those pages before beginning the multivariate tests. One was to avoid having two calls to action on one page; another was to make sure the call to action appeared above the fold on that landing page.
“These tweaks are things you can do ahead of the full-blown testing,” Colacurcio said. “There are a lot of basic things marketers aren’t aware of, and Widemile said out of the gate, ‘Before we create our baseline, here are a few things to change.’ ”
Widemile tested six different factors on the landing pages. The first was the body copy, the quick description of the product. The second was the call to action (the actual wording on the button or link). Third was the text beneath the call-to-action button. Fourth was the caption underneath a screen shot of the product on the landing page, and fifth was testing an analyst quote versus a customer quote. The final factor tested was text versus graphics for instructions on how to download templates.
“At the end of the day, there were 1,024 different combinations tested,” Colacurcio said.
The extensive testing revealed that the analyst quote won “hands down.” “Which shocked me—I was betting on the customer,” Colacurcio said.
Smartsheet.com used the detailed test results to refine the 40 landing pages. Conversion rates rocketed even higher, to between 10% and 12% on average across those five tested sites.
“We got over twice the number of conversions on the same budget,” Colacurcio said.
Based on that success, Smartsheet.com rolled out those refinements to the landing pages. Across all 40, the average conversion rate was 7% to 10%.
“Ideally, the thing to do is to test every single page, but if you can’t do that and you simply take the learning from the limited tests, you will still see great results,” Colacurcio said. “We saw the best results on the pages we tested specifically, but we were able to get a good ROI by expanding that to the rest of the pages.”