Can We Measure Visitor Attention?



Last month, comScore changed from measuring page views to their new “visits” metric, designed to better measure visitor engagement. Or, as they put it, “visits” measures “the number of times a unique person accesses content within a Web entity with breaks between access of at least 30 minutes, is a way of measuring the frequency with which a person views content, thereby illustrating a key component of user engagement.”

Nielsen//NetRatings is not one to be left behind. The The Wall Street Journal (sub req) reports on NTRT’s new metric: “Nielsen/NetRatings, in June will release what it calls ‘time-spent’ data and stop issuing its rankings by page views.”

Bryan Eisenberg asks the obvious questions:

  • Do you ever open up a browser and get distracted by a phone call, a meeting, your kids, or an instant message? Will the time the “page” is open be counted there as well?
  • Have these folks never heard of tabbed browsing? Stop reading for a moment. How many tabs or browsers do you have open at the moment. More than one, most likely. People keep 6 or 7 tabs opened and would all of them count as time spent from the moment that tab was opened? That’s a clear indication that someone is engaged, huh?

And here’s one more obvious question: what do they mean by “time-spent” data? I mean, seriously, they can’t really just be measuring the amount of time a page is open on a browser window… right?

  • http://www.prezzatech.com Brian

    I agree. I think anything longer than 30 minutes on most sites is either a video site or a game site.

  • http://www.scentedcandlesclub.com AnnaLaura Brown

    I think this is a great idea but difficult to implement. I would love to know more about what my site visitors are doing and reading.

  • rick gregory

    This is the standard definition of a unique visit. The other way to do it is to count each visitor once per day (or some other long interval) which guards against the “I got distracted for a bit” issue.

    The other way, of course, is to measure activity on a page, i.e. to actually measure activity versus having time proxy for activity.

  • Jordan McCollum

    The comScore visits definition isn’t at issue; the undefined but unfortunately named “time spent” metric is what I’m questioning.

  • rick gregory

    I know… hence my other suggestions. I think if we want to measure engagement we need to find a way to measure activity, not just whether the page is open or not. I know we can measure clicks… not sure about the time interval between clicks… that might tell a site owner something interesting esp if they combined it with user studies.

    The real wrench in this is, I think, the tabbed browsing issue you mention… the “I got distracted” stuff feels minor… tabbed browsing doesn’t

  • http://andybeard.eu/ Andy Beard

    They can actually measure attention.

    Javascript on a page afaik one registers what is happening on that page, and we have all seen the cool click tracking and mouse tracking that is possible. You can also track screen scrolling.

    Lots of things are possible, the question is whether they are actually doing it.

  • http://www.u-g-h.com Owen

    There’s lies, darned lies .. and statistics.

    The problem (or not) with webstats is that they can be twisted in so many ways to prove something that it’s hard to believe anything. “Time spent” usually means whatever is convenient at that time.

  • http://www.webmaster-money.org Webmaster Money

    Nice Idea, it would be nice to see such measurement. But I think you can see in some statistics program atleast entry page and exit page and I also saw stats of user which pages he saw. This looks similar to this idea.
    I think that it could slow down the page with high volume of visits.