Quantcast

«

»

Apr 03 2014

Clickbait

One thing that I have learned by blogging is that I am absolutely hopeless at predicting what posts will garner attention. In fact, I seem to have a reverse sense, in that posts that I think are interesting sink without a trace while those that I think are uncontroversial generate quite a bit of buzz.

While having many readers is gratifying, since I mainly write in order to clarify my thoughts and express my views, I do not actively seek to write in a way to increase readership. But many online sites depend on the size of their readership for their income and for them, increasing readership is critical. The catch is that the yardstick that has come to be used to measure this is the ‘clickthrough’.

But as Tony Haile writes, we use a faulty measure for the popularity of a site.

In 1994, a former direct mail marketer called Ken McCarthy came up with the clickthrough as the measure of ad performance on the web. From that moment on, the click became the defining action of advertising on the web. he click’s natural dominance built huge companies like Google and promised a whole new world for advertising where ads could be directly tied to consumer action.

However, the click had some unfortunate side effects. It flooded the web with spam, linkbait, painful design and tricks that treated users like lab rats. Where TV asked for your undivided attention, the web didn’t care as long as you went click, click, click.

This spurred a huge effort to entice people to click on links. Some people are really good at getting people to click. The catch is that 55% of users spend less than 15 seconds on a page they reached. The number drops to 33% for pages that have actual articles. Haile’s team looked at the kinds of pages that people spent more time on and the results are somewhat encouraging in that people seem to spend time on more worthwhile pages..

Articles that were clicked on and engaged with tended to be actual news. In August, the best performers were Obamacare, Edward Snowden, Syria and George Zimmerman, while in January the debates around Woody Allen and Richard Sherman dominated.

The most clicked on but least deeply engaged-with articles had topics that were more generic. In August, the worst performers included Top, Best, Biggest, Fictional etc while in January the worst performers included Hairstyles, Positions, Nude and, for some reason, Virginia. That’s data for you.

All the topics above got roughly the same amount of traffic, but the best performers captured approximately 5 times the attention of the worst performers.

But until we can find a way to better measure actual engagement, we are going to continue to be flooded with clickbait pages.

So in the near future expect a post from me titled Virginia’s biggest nude hairstyles top awards for best fictional position. It can’t lose.

7 comments

Skip to comment form

  1. 1
    Dunc

    The great thing about clickthrough, and very probably the main reason it caught on as a metric, is that it’s really easy to measure.

    Oh, and you appear to have dropped a close italic tag…

  2. 2
    John Horstman

    In an upcoming article, Mano tells you One Weird Trick from a Mom in Delaware that you can use to Increase Site Traffic and Engagement 537%!!!

  3. 3
    Mano Singham

    @Dunc,

    Thanks. I’ve corrected it.

  4. 4
    busterggi

    Dammit, don’t leave bigfoot out of your post title – Sharon hill over at Doubtful News says he’s a click-magnet.

  5. 5
    Trickster Goddess

    I work at a small community newspaper. We have had a web presence for a number of years but only now are we moving to place ads on the website. Last week we met with the consultant we hired and he was going on about having the ads link to the advertiser’s web site and counting clickthroughs as a metric to show how successful the web ads are to other potential advertisers.

    I pointed out that it isn’t really a useful data point since most of our advertisers are local businesses, many of which don’t even have a website. Besides, people who see an ad for say, a local barber, aren’t going to bother going to his website, they are just going to go down to the shop for a cut. And for the barber, customers coming through his door is much more meaningful metric than visitors to his website.

    For over a hundred years businesses have seen value in displaying ads before the web was even invented. Web ads still have that same value, and the best metric is probably still “impressions” or views.

    Unless you are conducting actual sales through your site, the only thing a potential customer is going to get from clicking on your ad is a chance to to view a bigger ad. A company like Coca-Cola cares much less about you clicking through to their website than having their ad influence your choice of beverage on your lunch break.

  6. 6
    rq

    And the introduction to your post, Mano, should end with: “… And you won’t believe what happens next!!”

  7. 7
    Darragh McCurragh

    Actually there are various studies on the subject. After all, companies who invest huge sums in their campaigns and websites are not primarily interested in clicks but in “conversion rate” (if they’re e.g. an online shop. Others who “sell” information are interested in long viewer time, i.e. low bounce rates. Those studies paint a different picture: articles that are shared on social networks e.g. are rather longer articles while even “popular” articles get shared less often. And Google will rank a pertinent longer article higher than a shorter one on the same subject (which has many reasons probably, like more lines meaning more keyword collocations etc.). As in real life, value will prevail in the long run. Still, average (average, median, whatever) readership will decrease, as the amount of potential readers is fixed while the amount of readers-turned authors is still rising.

Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>