Digital Media Buzz > Using Analytics For Your Next Redesign

Using Analytics For Your Next Redesign

By Eric Bishop

It was one of those minor design changes that you really wouldn’t expect to move mountains. Pedro Sostre and his team at the design firm Sostre & Associates simply added an extra link to the horizontal navigation of a clients’ website. But the move killed performance about 30 percent. What happened?

Web analytics provided the answer. Sostre looked at heat maps showing visualizations of users’ clicks on the homepage. He discovered that users had all but stopped using the search function, which had been pushed farther over by the new navigation item. “The fact that the search bar moved over 150 pixels caused the usage of it to drop by 80 percent,” Sostre says.

Analytics — by providing information about how people arrive at a website, what they do there and whether they convert on the site’s goals — can illuminate problem areas and help webmasters improve them. The data are particularly helpful during a redesign, when information about the effectiveness of site components can guide the revamping process. “If you’re spending money on a website redesign and you don’t have analytics, it’s like taking a shot in the dark,” says Postre, who is the author of “Web Analytics for Dummies.”

Many businesses, though, continue to fire away at redesign projects without the guiding light of analytics. Cost is certainly no excuse — Google Analytics and several other leading tools are free. Lack of education, with regard to the benefit of analytics and the right way to use them, is a more likely reason for the deficiency. “A very small portion of businesses actually know how to make actionable decisions based on that data,” Postre says. For that reason, he suggests having an outside firm evaluate the analytics and make recommendations over the course of the redesign.

Google Analytics, for its part, has taken pains to simplify the process of retrieving and interpreting metrics. “We’ve done a lot of work on our end to redesign the interface to be much more user-friendly,” says Brett Crosby, group product marketing manager for analytics at Google. “Even novice users should be able to go in here and understand what’s going on without too much explanation.”

Crosby recommends baby steps. “The first thing I would do is try not to be overwhelmed,” he says. “It’s worth people putting a stake in the ground and saying I want to see this number go up. Pick your metrics that you care about and monitor those over time. Instead of trying to do everything, start with something small. Work on that. As you do work on that problem, the rest of it will start to unlock and become much more comprehensible and actionable.”

Many webmasters can glean key insights from funnel reports, which detail clicking patterns over a specific conversion path that spans several pages, such as the steps of making an online purchase. If a high percentage of users ditch the path when they get to a particular page, for example, that’s a signal that the page’s design or content could use a tweak.

Looking at how users are accessing the site — whether directly, from search engines or from other links — can also guide decision making. “It paints a better picture of what your users’ interests are,” Postre says.

Whatever metrics you draw upon, it’s important to make nuanced inferences from them. Higher average time on site is generally a good thing, Crosby says, except when users are spending way more time than they should be. Same thing with page views — more is usually better, except in the face of fewer conversions. And so on.

It’s also vital to draw accurate connections between site changes and usage statistics, which might be easier to do under an incremental redesign. Postre cites the added navigation link as an example where misinterpretation of the data might have been costly.

“Somebody could have looked at that issue and said ‘People have kind of stopped using the search box now. Is that because our search results are not displaying the right information?’” Postre says. “You could spend $50,000 tweaking the search engine algorithm to get you better results, but it wouldn’t matter because that wasn’t the case.”

Eric Bishop is a Brooklyn-based writer and web producer who recently worked as a nighttime news assistant for NYTimes.com. More of his writing is available at bishopsclips.blogspot.com.