Digital Product Design and Development

New York, NY

Screen Time Part 3

This is a short post to close out this series. See Part 1 and Part 2.

All standard web metrics are based on the idea of hits. A user requests something, the server logs a hit, and then the hits are assembled into metrics like pageviews, visits, unique visitors, pageviews-per-visit, visits-per-user, and visit duration.

Most web advertising is also based on hits. CPM, CPC, and CPA are all costs per hit.

This works well because hits are easy to define and easy to count. They’re easy to conceptualize, to talk about, to transact on. Hits are things. We like things. We can hold on to them. We can measure them, compare them.

Hit-based metrics aren’t perfect but they do a pretty good job with the stuff they’re supposed to do.

I don’t know if elapsed time is the exact opposite of hits, but at least in the context of web analytics, I think of it as an opposite mode of measurement. Instead of counting events we’re thinking about the space in between events.

Time-based metrics have been around forever but they’ve had a second-class status because (1) crude tracking methods have made the accuracy of the data questionable and (2) time-based metrics aren’t used to drive the important business conversations.

Smarter technology and improved data collection has solved the first problem. For years, more accurate time-based data has been achievable via analytics services like Chartbeat and Clicky.

The second problem is not really a problem but an opportunity. Should time-based metrics have a seat at the adults’ table? Medium thinks so. Upworthy does, too.

This isn’t about the ‘ole “Average Time on Site”.  We can get much more granular, measuring not just how much time people spend actively engaged, but where exactly they spend that time. We can measure how much time things spend on-screen.

As I mentioned in the first post, when I started looking at time data this way I was amazed at the total aggregate values as compared with the commonly referenced averages). Collectively, people spend a lot of time on the web, and an “Average Visit Duration: 2 min” datapoint doesn’t convey that.

In media, a recurring and crucial question is how do publishers define value for advertisers? I think there’s something interesting about framing this question in terms of time and brand exposure. “This week your brand was on screen in front of our audience for 2,500 hours” (CPH™, oh yeah).

Candy is dandy but data is quicker

To explore this idea some more I made a small client side tool called Screentime. It lets you define regions of the page or small elements like ad units, and record how much time each one spends on-screen. It’s in beta but ready to be played with if you’re interested in learning more about how your site activity translates to screen time.