Can you use the amount of time a user spends on a website to measure attention? The NextStage CRO thinks not.
A message board that I follow recently went through a series of posts about "attention statistics." I came too late to those messages to respond there, but I believe that the iMedia readers might find a discussion of the topic worthwhile.
The concept of "attention statistics", as described by Richard McManus, measures attention as "the total time spent on a site as a percentage of the total time spent online by all U.S. internet users." One factor that can affect attention is "velocity: the relative change in daily attention; velocity is used to determine the relative growth of a website compared to other sites."
Attention as a time-on-site metric confused me. At NextStage -- and I apologize ahead of time if my talk about our work seems at all like a sales pitch: that's not my goal -- we have a different take on "attention" and two ways to use this kind of information.
What is "attention"?
First, NextStage does make use of time-on-site in some of its reports. It's a valid measurement for certain things, but I don't think it's a valid measure of attention.
If what I'm about to write seems technical, bear with me or simply skip to the next section.
Both NextStage's and others' research indicates that "attention" is a mixture of intentional awareness, focused cognition, meaningful noise factors, intentional and unintentional attention factors, as well as a bunch of related psycho-behavioral markers. Some of NextStage's original proof-of-concept tests for our Evolution Technology involved Sensory to Mapping, Teleology and Time-Normalization Studies, all of which make use of Time-on-Task metrics.
But "time-on-task" isn't the same as "time-on-site," None of our "attention" reports -- such as Level-of-Interest (what are visitors paying the most attention to?) and Visitor Loss-of-Focus (where did visitors stop paying attention?) -- make use of time-specific measurements. And this is because things like intentional awareness, focused cognition, et cetera, do not involve a time-constraint on what's being measured.
Examples of Level-of-Interest and Loss-of-Focus occur when most people drive familiar roads, such as going to work and back or the store and back. Most people, once they've learned the roads to a routine destination stop paying attention (or what's called intentional attention) to the drive.
Many people no longer even focus (intentional awareness) on what they're doing when they're driving. Everything is happening at an unintentional attention level, only coming into conscious awareness when the need arises. The only time you're directly aware of the drive -- i.e., thinking about what's going on or focused cognition -- is when something out of the ordinary happens. There's an accident, a police car goes zooming past, or when you have to go to another office and are looking for landmarks, street signs, exits, so on and so forth. When these things happen you've gone back to paying attention to what's going on; otherwise you're not going to bother.
Another example of time-on-task being a questionable metric for attention comes from my turning on the TV most Sunday afternoons then taking a nap. I definitely "spent time" on the show, but trust me, I wasn't even aware of what was on. Anybody want to claim I was paying attention?
Attention is focused cognition and intentional internalization of information
There are a bunch of easily captured physiologic markers that occur when someone is interested in something on a website, online video, sound file or marketing brochure.
At NextStage, we call two of these things Level-of-Interest and Visitor-Loss-of-Focus, and we've been studying them for our clients for years.
First I'll share how this information serves in both media design and strategic planning. Later I'll share how it can be applied to blogs.
Design and planning: Emetrics Summit case study
NextStage has been monitoring the various Emetrics Summit sites for a few months and supplying the Emetrics Summit folks with suggestions for improving traction on their sites.
Often these suggestions come from the Visitor-Loss-of-Focus report. The information helps Emetrics Summit organizers modify pages where visitors lost interest.
And note that "losing interest" is not the same as leaving the site or dropping off: those are different than the places where visitors lose interest.
Fixing these pages keeps visitors on the site, branding them and keeping them in the sales funnel.
Likewise, the Level-of-Interest reports closely align with what's going on in the greater Emetrics Summit world, not just on the websites. This real-world to virtual-world congruence is shown in the following chart:
Based on these and other congruences -- none that NextStage was aware of until we shared our findings with Emetrics Summit organizers -- NextStage will be advising Emetrics Summit organizers on which geographic locations are showing high levels of interest. The end goal is to determine where future Emetrics Summits should take place.
When real-world and virtual-world activity do not converge we know there's a fixable problem with a site. In addition, these metrics predict what should be happening in the real-world. And, again, when they don't, then we know there's a fixable problem with a site.