In a previous blog, we discussed preliminary, site-level data around the LIVESTRONG.org redesign and launch. With three months of data now available, we revisited the site
hoping to better answer the question: how does the new site fulfill its goals as opposed to the old site? An A/B test would be the best, but in the absence of one, and with limited insight into the old site’s on-site activities, we decided to revisit some of the metrics we pulled following launch and pay special attention to bounce rate of new visitors who accessed the site via search. Here’s why we did it, how we did it, and what we found …in three easy steps!
Step 1: Identify Company and Web Site Goals
As an organization, LIVESTRONG seeks to help anyone whose life is affected by cancer. So at a very high level, a successful visit to LIVESTRONG.org could include:
- Effectively accessing information about cancer, cancer support, or available resources.
- Learning how to donate time or resources to the LIVESTRONG cause or about other ways to get involved in sponsored events.
Based on that, an improved Web site would not only have to better help persons who already use the site, but it would have to indicate a propensity for helping persons who didn’t necessarily know where to look in the first place.
Step 2: Identify Metric(s) that Speak to Those Goals
To find the information they need, people undoubtedly need to look at more than one page and spend a little time on the site. For that reason, we decided to update the time-on-site and bounce rate metrics we examined at launch to see if the positive trends continued through July.
To gauge the new site’s ability to assist new community members, we turned to the bounce rate of new visitors who arrived via natural search. We focused on this segment because there’s a good chance those visitors weren’t necessarily looking for the LIVESTRONG domain when they hit site, even though they likely constitute people LIVESTRONG hopes to assist.
Step 3: Pull the Data and Compare
While pages per visit have remained comparable since Jan 2010 (between 2-3 pages per visit), it appears people elect to spend more time on the site.
- Bounce rate has continued to decrease month-over-month since launch — ultimately dropping from 63.67% in March 2010 to 44.56% in July 2010.
- For the three months following launch (May 2010-July 2010), visits to the new site have averaged 4:18. From January 2010 to April 2010, the old site averaged only 1:58 per visit.
People are finding more engaging content more efficiently, and when they land on the new site, they’re choosing to view more content more often than with the previous design.
We also found that as new visitors hit the domain through search, they’re electing to stay more often as well — even more so than other types of visitors.
- Bounce rate among new visitors arriving at the site through natural search has continued to decrease month-over-month since launch — ultimately dropping from 62.94% in March 2010 to 30.74% in July 2010!
This is great news from a design perspective, since these folks likely represent an opportunity to expand LIVESTRONG’s community.