In a time when Google and Facebook dominate web traffic and publishers are struggling to keep pace, pageviews have never been more important. But how can news organizations compete with the duopoly?

Apparently by going back to basics.

According to a study conducted by the Engaging News Project — an initiative of the Moody College of Communication at the University of Texas, Austin — redesigning a website's homepage could increase pageviews and the amount of time users spend on the page. But the results were mixed.

The study was conducted in conjunction with the launch of two major news organizations' redesigned homepages, one Canadian and one American. Around the same time, nearly 1,000 Canadians and 1,000 Americans participated in online experiments, in which they were randomly shown either the old homepage or the updated one, and then asked questions about their perception and recall.

Here's what researchers Natalie "Talia" Jomini Stroud, Emily Van Duyn, Alexis Alizor and Cameron Lang found:

  • For the Canadian redesign, pageviews and time-on-page were higher on the new site compared to the old site.
  • For the U.S. redesign, the bounce rate increased, and the average time per visit and the number scrolling halfway down the page declined on the new site compared to the old site.
  • For the Canadian site, article recall was higher for the new site than the old site. There were no significant differences in article recall between the old and new U.S. sites.
  • There were differences in which articles were recalled on both the U.S. and Canadian sites.
  • Some articles were better recalled on the new sites, and others better recalled on the old sites.
  • Differences in recall corresponded with the presence of images and where the articles were placed on the page.
  • Differences appeared across the old and new sites in both countries regarding how people rated the site, and the features they liked most and least.

"We don’t know the exact reason that the Canadian news site saw an improvement in their metrics while the U.S. site did not because multiple elements changed," said Stroud, director of the Engaging News Project, in a press release sent to Poynter. “One possibility is that the new Canadian site incorporated more photos on the part of the site that appeared ‘above the fold’ while the U.S. site incorporated fewer images.”

Despite the mixed study results, researchers definitively found that articles with less scrolling, a photo and located in left-most columns were better-recalled. And while people visiting the websites during the experiments were generally less engaged than those visiting on their own, the results from both sets of participants in terms of pageviews and time spent between the old and new websites were comparable.

The bottom line: homepage design plays a key role in how users interact with a site.

"Different website designs can elicit different user experiences and influence what people recall," researchers concluded in the study. "Site design can affect metrics like pageviews and time on site, but results can vary."

The latest study is a follow-up to another Emerging News Project study from 2015, which found that people read more on sites with contemporary designs compared to those with traditional newspaper layouts. Taken in tandem, both studies show that redesigning a news organization's homepage to be faster, more visual and more user-friendly could boost the amount of time people spend on the site, while also increasing the number of articles they read and remember.

But given the most recent data from the American news organization, the quality of the results depends on the outlet. Emerging News Project researchers recommend that newsrooms looking to redesign their site conduct experiments first.

“Our results show that an online experiment can pick up on many of the same signals as a full launch of a site redesign,” said Van Duyn, research associate for the Engaging News Project, in the press release. “We believe that doing an online experiment could provide news organizations with a relatively inexpensive way to test out a redesign before a full launch.”