The business case for dwelling on dwell time

dwell time

I’m a big fan of the New York Times. Not only do they lead the way with the best in American journalism and commentary, they also are innovators in their approach to marketing articles to get them to move.

And as you know by now, my strong view is that we must get our content to move. In fact the economic value of all that content you’re producing is exactly zero … unless people see it and share it.

So a recent post by the New York Times that reported its most popular posts of 2015 by dwell time got my attention.

Dwell time is the amount of time people spend viewing your content. This measure is particularly noteworthy in today’s marketing environment because we know that Facebook has changed their EdgeRank algorithm to acknowledge this metric.

Like most people, I have historically judged a success of a piece of content by its page views. Naturally more people viewing my content would be a sign that it is successfully “moving,” right?

But if the New York Times is judging success by dwell time and Facebook is rewarding it, perhaps we need to reconsider how we are measuring content marketing success.

A metric revelation

Naturally I wanted to see what would happen if I compared my most-viewed blog posts of 2015 with the posts that people spent the most time viewing. Here is the comparison of the top five posts by views and dwell time:

dwell time

Well. My most popular posts by page views versus dwell time is mutually exclusive. I was kind of hoping for some overlap, but no such luck!

This is a small sample size but at least looking at the top five most-viewed posts for 2015, none of them also made the top five in dwell time.

The next natural question is, how long are these posts? Do people spend the most amount of time on the longest blog posts? That would make sense, right?

Well perhaps nothing about this investigation makes sense …

dwell time

This chart shows that four of the six shortest posts received the most dwell time from my readers. Huh?

Further, three out of the four longest posts did not make the top five list for dwell time (the red bars). Are people just skimming the long stuff?

Unraveling dwell time

Now that I thoroughly confused you, let’s review what we can conclude about dwell time.

  1. Dwell time is probably the most important metric you’re not measuring. It’s a key measure today because Facebook (and presumably the other social platforms) rewards content that attracts reader attention for a longer time.
  2. Dwell time is a signal that people are actually consuming your content, which is a leading indicator of advocacy.
  3. This is a small sample size … too small to make over-arching conclusions … but creating longer posts may not necessarily have an effect on how much people read them!

It might be worth doing your own content comparison. This is easily accomplished by going into Google Analytics and sorting your pages of content by clicking on both page views and average time on page:

dwell time

The best content measure would be something that combines content movement with content viewing — but more on that in an upcoming post!

Your thoughts on the importance of dwell time?

mark schaefer

Illustration courtesy Flickr CC and Mike Licht

All posts

  • Loved this piece, and I agree, Mark, that this is a crucial metric. (I must also admit I’m not measuring it). I want to point of to you and your readers that “dwell time” and what google analytics reports as “time on page” are actually two different things. Google only reports time on page when there is a subsequent page view in that session. So if a reader spends 20 minutes digesting a post and then closes the browser tab that doesn’t get recorded. When I’ve done analysis in the past, this generally totally skews the data toward posts that have a decent dwell time but also link to other posts further down within the post.

    I know there are scripts to get Google analytics to actually measure dwell time, but I don’t have a link handy. The other metric to look at as a proxy is scroll position. I’ve found that easier to accurately report in google analytics and fairly insightful in the past.

    Thanks for sharing your data and your thoughts. This will definitely push us to start looking at these better engagement metrics.

  • I do look at bounce rate, which is like a poor man’s dwell time, I guess. This got me really thinking about what might cause people to stick around…could it be posts with a lot of comments, where people would spend time reading both the post and the comment stream? Did you look at that stat for the top 5?

  • Mark, very interesting, but your stats might be skewed because of email delivery. Look at your Feedblitz stats to see if that lends any other insight. For example, I spent a couple minutes reading your post in email, then went to write this comment. Before commenting, I checked the discussion to see if anyone else already made this point. There were only 2 comments. So my dwell time is only the time it’s taking to tap this out on my iPad. My dwell time would have been longer if there were more comments.

  • I was not aware of that Steve. Logically, it doesn’t make sense that Google wouldn’t credi that but now you have piqued my curiosity for further investigation.

  • Good notion Rosemary. Something to consider.

  • Good thinking Candyce. I can see how his could be skewed by a lot of things. At best we can see this as a proxy assuming that over time all those skews would work out evenly for ever post. Interesting idea that view time could increase for later readers who see more comments. Thanks for the contribution!

  • Mark,

    I do agree with you about the mis-measurement of metrics that you often write about.

    Dwell time most certainly says more about the quality of content than page views.

    I also wonder about some of the other common measures of success that should be compared.

    I get the feeling that a lot of content is shared, liked, and retweeted that is not even read (or vetted). For example: my Buffer stats for twitter shares often show more likes and retweets that click throughs. And, I have had blog posts shared within seconds of posting with clearly no time for reading.

    The problem is that so many declare these basic interactions of a like or a retweet as engagement. Can engagement really happen when it only goes one way?

    Authentic metrics are needed to truly measure success in social media. You seem to have the best handle on this… we need another book from you.


  • I have a new post coming out soon that addresses this very topic Denny. I have a new idea for a way to approach measurement that I think will resonate with people.

  • Please share what you learn on that. Steve’s comment was a surprise to me too.

  • I’m really looking forward to this!

  • Pingback: The One Detail These 15 Content Marketing Thought Leaders Absolutely Want You to Know()

  • Pingback: Jumpwire Social Media Briefing - August 8 - Jumpwire Media Agency()

The Marketing Companion Podcast

Why not tune into the world’s most entertaining marketing podcast that I co-host with Tom Webster.

View details

Let's plot a strategy together

Want to solve big marketing problems for a little bit of money? Sign up for an hour of Mark’s time and put your business on the fast-track.

View details


Send this to a friend