You’re a proponent of quality over clickbait. When we talk about quality, usually it’s a judgment call. So how do you get around that and turn it into an objective measure?
It’s always very dangerous when you talk about quality. Quality is incredibly difficult to measure.
What we can say is that in general a visitor’s default behaviour is not to read every single word that is written on a web page. A visitor’s default behaviour is to leave, and they will use almost any excuse to do so.
With that in mind, if you are able to hold someone’s attention – and bear in mind that 55 percent of all page views get 15 seconds of engagement – you must be doing something right.
What about “brand”? Do you think it can be measured in a meaningful way?
One of the great siren songs of the web is the promise that we’d be able to measure so much more.
One of the dangers for brand advertisers is that they’re so keen to bring some clarity and measurements to the brand space that they jump on to direct response measures and start using them as a measure of success when they’re really not.
Rather than trying to apply every metric we possibly can whether appropriate or not, we should understand what we can do, and then allow creative to push the message through.
One of the great siren songs of the web is the promise that we’d be able to measure so much more. Tweet
Chartbeat’s real-time analytics tools give publishers the ability to know the impact of every editorial decision. Plenty of editors rejoice at this power, while it probably makes others feel nauseous. Do you think there’s room for editorial judgment in the publishing world anymore?
What’s key for every editor and writer to remember is that our job is not just to write the important stories, but also to communicate them.
If you have written an important story about Afghanistan and people aren’t reading it, I’m not suggesting that you drop the story and start writing about cats. I’m suggesting that you work out how to get that message across.
Data from services like Chartbeat should be followed just like the editorial feedback from a managing editor. It’s another circle in the feedback loop that can help journalists communicate their stories better.
When you start looking at data beyond page views and clicks and actually look at attention and so forth, you can find out just how much people are reading your stuff.
Sometimes that can make you really uncomfortable, but it’s the truth and it’s useful to embrace it.
There is a push to embrace time and attention as markers of value on the web. But as Evan Williams, CEO of Medium, recently wrote, “Advertisers don’t really want your time – they want to make an impression on your mind.” How do you reconcile this tension between time and attention?
There is peer-reviewed research showing that for brand advertising, there are two things that make a significant difference in terms of recall and recognition. The first is the quality of the creative and the second is the amount of time it spends in front of someone’s face.
In March of last year IAB made the standard that a viewable impression happens when 50 percent of the ad is seen for more than one second. What they’re saying is that for the very first time, there is a quantitative difference between two ad impressions on the same page based on user behaviour.
There are only two things that a publisher can effectively control. One is the audience that comes to their site. Second is the amount of attention that is captured.
Once you control for audience and attention, every other metric that has been thrown against publishers are measures of creative. Whether it’s click-through rates, dwell time, or interaction with the ads, those are all measures of creative and the creative agency should have responsibility, not the publisher. That’s a 20-year mistake.
And then there’s programmatic, which is getting bigger and bigger and which runs on CPMs. This doesn’t sound like good news for publishers wishing to find other ways to value their ad space, does it?
Programmatic is running on CPM right now because that’s what the ad world runs on. It’s a function of logic not a function of anything intrinsic to programmatic.
I actually think some of the most interesting stuff coming out in the next couple of years will be the application of programmatic to brand metrics rather than just straight direct response methods.
It’s going to involve more complicated math, it’s going to involve much more interesting logic but I think there’s tremendous potential in that space.
Let’s talk about work culture. Designers, data scientists, developers and lots of engineers work at Chartbeat, but it’s also a relatively small team. How do you foster ideas good enough to compete with the Googles of the world?
One of the things we really encourage is for people to speak up. We’re still small, and when you’re only one hundred people competing against huge companies, you need to leverage the creativity and intellect of every member of the team.
So we give them significant periods of time to be creative, to try crazy things whether they relate to our products or not.
When you allow a certain amount of redundancy or experimentation, you uncover brand new ideas and products that you never would have thought about, and that’s the key.