22

Jan

By Dennis R. Mortensen
22 comments
News Media that use Galleries to increase Page-views don’t increase Time-spent (Attention)

It is not uncommon to see News Media Publishers drive un-segmented page views as one of their primary Key Performance Indicators (KPI’s) – And one of the biggest page view drivers for online News Media today are the common use of Galleries (slideshows if you will). I would argue that those additional page views, however massive they are, don’t deliver the same value to the advertiser, as compared to the core product (the news article) – and ultimately isn’t really what the user/reader wants.

I studied the usage patterns of one of the biggest News Media publishers in the US as part of an optimization dialogue and created the following four News Media content segments:

  • Galleries
  • News articles
  • Front pages
  • Other

Giving me a much better understanding of where the bulk of the page views and thus the advertising inventory is generated. There is no surprise in the below bar chart that visualize the total number of page views per news media content segment.

news-media-page-views

Galleries make up for almost 60% of all page views, which is a dramatic number. I believe it is dramatic, because you don’t have to conduct to many studies or interviews, to find that the value of a gallery page view, is probably not the same as news article page view. I was therefore interested in telling a different story with the same data.

I summed up the total time spent on each of the content segments using a simple time spent metric. Whether you believe that total time spent on news media content segments is identical to reader attention, probably doesn’t matter too much, as I am sure you would agree that, if not, it is at least a decent proxy metric.

There might be a positive surprise in the below visualization, which quite clearly shows that galleries might have the most page views in total, but that the core news media article product (and the front page it is promoted on) is where the reader spend most his time, and I would argue his attention.

news-media-time-spent

It is therefore somewhat sad to see quality news articles getting thrown into ad exchanges and sold alongside galleries. I was therefore positively surprised and happy to see that Nick Denton and his Gawker Media announce their move away from measuring success on page-views. Starting 2010 they will measure and compensate success on Unique visitors. I’m no direct fan of one metric over the other and am not sure gawker has the ability to accurately calculate a true unique visitor. I am however a fan of not valuing every page view the same, simply because, any decent publisher knows how to increase page views without adding any real value to their publication (yes, I am talking about galleries).

- OR as Nick said it “An item which gets picked up and draws in new visitors is worth more than a catnip slideshow that our existing readers can’t help but click upon.”

There is some really good commentary on the subject on The Nieman Journalism Lab’s website as well.

Cheers :-)
/ Dennis (@dennismortensen)

  • http://www.emerkirrane.com Emer Kirrane

    Hi Dennis – great post!
    I agree – galleries are part of instant gratification and an add-on and tend to be “flicked through”. Nice but no spice, and all that.
    Personally, I tend to think that all page views should have a segmented weight. Pages are not all equal (generally) and simply counting views is pointless without context.
    Very interesting to see Gawker move towards Unique Visitors – quite forward-thinking to look at attracting visitors rather than just clicks per visit, essentially.
    Cheers, E

  • http://visualrevenue.com/blog Dennis R. Mortensen

    Hi Emer,

    >>Pages are not all equal (generally) and simply counting views is pointless without context.

    Indeed. AND sometimes it seems like Publisher’s keep at it, simply to please advertisers – which is somewhat misguided.

    Cheers.. and have a great weekend

    d. :-)

  • Matt S.

    Hi Dennis,

    Interesting…one question… I’m wondering if you segmented “article” visits w/gallery interaction vs. “article” w/o gallery interaction? I would imagine that the duration times for the gallery folks is higher and that “Gallery” interaction may be a good indicator of high engagement.

    thanks,
    matt

  • http://visualrevenue.com/blog Dennis R. Mortensen

    >>I’m wondering if you segmented “article” visits w/gallery interaction vs. “article” w/o gallery interaction? I would imagine that the duration times for the gallery folks is higher and that “Gallery” interaction may be a good indicator of high engagement.

    Great question Matt. And I actually did take a look at that, and the results was actually quite the opposite (perhaps I’ll do a post with that data as well). One thing galleries does well though, partly in line with your idea of them not being valuable in and by themselves, but as drivers for something else, is as appetizers for aggregators. So the traffic volume (new visitors) generated from aggregators picking up your catnip slideshow (using Nick’s wording) is something you certainly can use galleries for – it just seems like those visitors are neither loyal nor really engaged.

    I know that it shines through that I am personally not a fan of one slideshow after the other. But I am doing my best to be objective. Help me. :-)

    Cheers
    d.

  • Matthieu

    Interesting !
    A good idea would be to measure if the advertising is actually seen, which is technically very very easy, and already done by Alenty : http://www.mondaynote.com/2009/05/24/measuring-time-spent-on-a-web-page/
    But I think agencies wouldn’t be happy : advertisers would want to pay for ‘seen ads’ and not ‘displayed ads’

  • http://visualrevenue.com/blog Dennis R. Mortensen

    Hi Matthieu

    >>advertisers would want to pay for ‘seen ads’ and not ‘displayed ads’

    Exactly my point!

    With the data set above, only 0.27 seconds (322K hours / 72M PV’s) are spent on every page (slide) – which in my opinion, is indicative of insignificant engagement to the extent of “NOT having seen the ad“. On the contrary, News Articles have almost a 700% higher attention rate (again according to the above data set).

    Cheers
    d.

  • http://www.golfbusinessmonitor.typepad.com Miklós Breitner

    Hi Dennis,

    You are only talk about galleries, but how about videos? and video sharing portals?

    cheers,

    Miklós

  • http://visualrevenue.com/blog Dennis R. Mortensen

    Hey Miklós

    Good to hear from you. I didn’t do any studies on the use of video’s – but I would suspect a result that it is quite different. And that to the positive side.

    I am not sure I follow you in regards to sharing portals, that’s an off-site activity and not really related to this debate I think.

    Cheers (-well good night I believe, back in Buda)
    d.

  • http://chattarati.com John Hawbaker

    As a web analyst—and an editor of a local news web site—I agree with you here. Page views may calculate how much CPM revenue you bring in, but they don’t provide any guiding insights.

    For a local site in particular, other questions are more interesting. Are our visitors primarily local, or are they coming to the site from Reddit because of a controversial story? What percentage of our visitors are loyal, coming back every week or every day? And how are visitors who don’t bounce consuming content and spending time on the site?

  • http://visualrevenue.com/blog Dennis R. Mortensen

    John,

    spot on! – and very much along the lines of what I am thinking.
    Thanks for commenting.

    nb. Which news media are you working for?

    Cheers
    d :-)

  • http://jorge-delgado.com/blog Jorge Delgado

    Great post…and amazing statictics…

    Thanks for sharing,
    Jorge

  • http://www.contentmetrics.de nic

    Hi Dennis,

    a few years back I worked for one of the big portal sites. At least some of the guys buying our ad space were very, very aware of this and paid a much lower price per AI on galleries. Some others just bought our X-AIs-on-sports-pages-package without looking too deep into it, though:-)

    On the other hand, some pictures have an enourmous effect on click-rates by touching people emotionally. Would be interesting to see, if your Time Spent metric correlates with ad-clicks (although I doubt that, but I like to be surprised).

    so long
    nic

  • http://visualrevenue.com/blog Dennis R. Mortensen

    Hey Nic,

    Thanks for commenting. AND it’s nice to hear that we have SMARTER advertisers out there. – Not a bad question at all, in regards to correlation in between ad-CTR and Galleries, so I am taking notes for future posts.

    n.b. I would love to hear any indication of the Gallery “Discount” ??

    Have a great weekend
    d. :-)

  • http://ethicminds.blogspot.com george naing

    Intuitively, we suspect that gallaries etc are more of a distraction than anything else. Now you provide some substantial proof.
    BTW, here I recommend your book. I also thank your here for your book

  • Kris G

    Dennis I’m a first time reader, and this is the first post I’ve read. It’s safe to say I will be a long time reader now.

    I’m an analyst for a major media company in Canada, and would agree with a lot that has been said. Video definately brings much more engagement, and with pre-roll video ads available, advertisers get great exposure and pay more for their value.

    Kris

  • http://visualrevenue.com/blog Dennis R. Mortensen

    Hey Kris,

    Thanks for the thumbs up mate. AND I agree that Video is not to be aligned with Galleries – at all – actually.

    d. :-)

  • http://charlotteobserver.com/dailyedit Gary O’Brien

    Dennis:

    I’m glad to see you addressing this question. I agree with a lot of what’s been said here, but I also believe what nic said above:

    “On the other hand, some pictures have an enourmous effect on click-rates by touching people emotionally. Would be interesting to see, if your Time Spent metric correlates with ad-clicks (although I doubt that, but I like to be surprised).”

    My experience is that when we have a particularly touching or emotionally compelling set of images that our time spent tends to go up.

    Keep up the good work – this is my first time here, I will be back.

    Gary

  • http://visualrevenue.com/blog Dennis R. Mortensen

    Hey Gary,

    Thanks for commenting. AND some very valid questions (observations). I honestly believe that news media in particular have been starved in regards to ACTUAL insights, so I certainly have more comments to make on the subject matter..

    One note (even though that’s probably not what you are implying). A Time Spent increase because of true interest (such as your scenario above) would in most cases, equal a fair VALUE increase, no matter the CTR rate. I do agree that for some ad systems or some ad sales teams, CTR might represent the advertising product value directly (or indirectly as a proxy for CPM value), but I think we’ll see that change.

    have a great weekend
    d. :-)

  • http://carsonhsmith.com Carson

    Very interesting analysis of a topic that, judging from the comments, seems near and dear to many web analysts. On my company’s site, slideshow visitors spent about 8% more minutes/visit on the site than article visitors. So on the surface level, it seems like a good selling point. However, looking at it in time spent in aggregate is really the ah ha moment. Love the way you presented the data here. At this time, it’s very difficult to convince editors to change direction when the advertisers don’t seem to care. I am working hard to emphasize unique visitors in all our reports, but until the marketing community decides to educate itself on web analytics (or when pigs fly), I can’t make a great business case to give up galleries.

    Just discovered your blog and look forward to following.

    Carson

  • http://visualrevenue.com/blog Dennis R. Mortensen

    Hi Carson,

    Well, if nothing else – YOU and I agree! :-)

    >>However, looking at it in time spent in aggregate is really the ah ha moment

    I’ve been doing presentations for News Media companies on this and given the right talk, proper data and the right visualization – most insiders certainly tend to get the point, but as long as advertisers are OK with the CPM not changing from Article to Galleries, they’ll continue. And from a business point of view, I actually don’t blame them.

    Have a wonderful weekend
    d.

  • Jon Tesser

    Dennis,

    I’m curious how you calculated Time Spent considering that it typically requires two time stamps (the entry page and a next page). A lot of users who come to a story from an outside source bounce after going to a next page, and therefore the Time Spent metric is basically useless for most articles.

    I’ve noticed that slideshows for our publication tend to outperform articles in terms of time spent (keeping in mind what I just said in the above paragraph), and are great performers across all segments (new, loyal, etc). Bounce rates are also way lower for slideshows than articles. We will continue to use them as a way to boost impression counts for advertisers that are ordering via CPM (as you mentioned above, this makes a lot of sense from a business perspective).

    Cheers
    Jon

  • http://visualrevenue.com/blog Dennis R. Mortensen

    Hi Jon,

    >>A lot of users who come to a story from an outside source bounce after going to a next page, and therefore the Time Spent metric is basically useless for most articles.

    That known fact, are more likely to provide an unfair advantage to Galleries, and even with that handicapped start, Articles win three times over (in the above example).

    >>I’ve noticed that slideshows for our publication tend to outperform articles in terms of time spent (keeping in mind what I just said in the above paragraph)

    Very interesting (I am seeing the opposite). I would love to see some of your data on that.

    cheers
    d. :-)