Non-transactional Analytics, Part 3

In this series, I’m looking at approaches to measuring the effectiveness of non-transactional Web sites, or Web sites with the primary purpose of providing information.

When a site doesn’t have a defined conversion point such as a sale, a subscription, or a registration, it’s often difficult to determine how well the site is doing. Quite often, site-centric measures such as visits and average visit duration are a blunt instrument in measuring success of these types of site. These measures lack context, so you’re left not knowing whether their value is good or bad.

When it comes to measuring success on non-transactional Web sites, more emphasis must be placed on listening to the customer’s voice. Part two of this series talks about the importance of putting a visitor satisfaction program in place as part of a framework to measure overall site performance.

This doesn’t mean Web analytics data don’t have a role to play in this framework; it’s just these types of metrics require some context to become meaningful. If a report says, “People who come to this Web site spend an average of six minutes there,” you’re left wondering whether that’s good or bad. Should you try to decrease this number or increase it? It’s difficult to answer these questions without context.

Many non-transactional Web sites, such as corporate portals, are designed to serve a number of different audiences, such as consumers, customers, job seekers, media, and investors. Each audience will have a different way of interacting with the site; it’ll have a different motivation and journey. Six minutes may be good for one group, as it suggests those users are consuming and engaging with the content. But for another group, six minutes may suggest they’re struggling with a particular process or aren’t able to find the content they seek. It’s key to provide the context for the metrics by segmenting visitors into different groups. Your ability to do this will, of course, depend on the Web analytics tool you use.

Context can also be provided by looking at the content, or site pages, in more detail. The publisher might have a tendency to create all the site content the same way. However, different pages have different jobs to do. At the simplest level, you can categorize content into three different types: inspirational, navigational, and informational.

Inspirational Pages

Inspirational pages are primarily landing pages where visitors start the visit. These pages should inspire visitors to make the next click. One key metric for these pages, then, is the bounce rate. The bounce rate is the proportion of all visitors who land on that page and exit the site without looking at any other pages. The lower the bounce rate, the more effective the page is in getting visitors to take the next step. The bounce rate isn’t always provided as a metric in Web analytic tools, so you may have to calculate it yourself by dividing the number of times this page was a “single-access page” by the number of times the page was an entry into the site.

Navigational Pages

Navigational pages direct the visitor to relevant site content. They may be category heading pages such as “News” or “Sports.” On these types of pages, you want to direct people to the content they’re looking for. Useful metrics here include CTR (define), that is, the proportion of visitors who click through to another page. Data visualization, such as click overlays, is useful here to see which links are clicked on. Many leading Web analytics tools offer data visualization.

Informational Pages

Informational pages provide the information or content the visitor is looking for. Here, you probably want evidence the content is consumed, such as by looking at the amount of time those visitors spend on this page and whether they download a document.

The measurement of non-transactional Web sites using Web analytics tools is about context. Context comes either from an understanding of what different types of visitors are doing on the site or from the role of the page itself. Broad-stroke metrics such as average visit duration don’t provide the insight necessary to truly understand the effectiveness of non-transactional Web sites.

Join us for the ClickZ Specifics: Analytics seminar on May 2 at the Hilton New York in New York City.

Related reading