DENNIS MORTENSEN: We can be considered the evolution of the Bloomberg Terminal for the newsroom. Any editor, whether at Forbes or at People Magazine or the Independent or The Telegraph, needs some sort decision support when it comes to what stories they should carry, and where to place it. For the most part, the traditional method of editing still depends on good editorial judgment, gut-feeling and what they believe is important. We're not here to overtake that process. We're here to participate in it. Think of us as the editor's best friend. We’re powering the data-driven newsroom.
How do your tools help editors make better decisions?
When news is breaking on social media and other news sites, it’s impossible to have a complete real-time look at what’s happening at any given time. When you look at what's trending on Twitter, Facebook, Google or the internet in general, editors receive a demand signal indicating what content is in demand. Then comes the question of course, whether you produce content to that demand.
The editor to some degree starts there. But the majority of the job tasks that they have is to look at the pool of content, which they have already produced. Think of the New York Daily News or the New York Times for example, which produce about 200 original stories every day. When they look at that pool of stories, they need to decide: "Which story goes where? For how long? We're not there anymore, which new story do I put in to that position?" That is called “homepage programming,” that is called “the job of the editor.”
By figuring out the taste of the audience on a Monday at 9am, for example, versus a Wednesday at midnight. It's certainly not the same. It's certainly not the same from week to week, because three months ago, users may have had a stronger political interest than they have today.
Although traditional publishers have been starting to embrace biddable ad sales to some extent, they are still a little wary about algorithms taking over their decision-making. How do you deal with editors who are concerned about the encroachment of data into their traditional territory?
When people talk about data, they envision some access to a plethora of graphs and charts. We don't deal with analysts. We deal with editors and we understand their needs and concerns. If I just throw up a set of charts and numbers, I'm not so sure that they'll be any wiser. What we try to come up with is a set of insights, which we have distilled for them, based on their objective. We can say, "Take that story which you wrote about Spain and their banking crisis, exactly 17 minutes ago and put that into folder picture one. If you are willing to do so, you stand to make an additional $172 dollars on top of what you're already making."
That is something that an editor is not confused about or threatened by.
And to make it work, to have credibility beyond just data points, we obviously need a very good understanding of their audience and their taste. We need to demonstrate a good understanding of the layout of that homepage and section pages. That's how, to use the term, Holistic Yield Management comes into play here, in that no piece of content can really be valued the same. For example, a national entertainment gallery doesn't yield the same value for the publisher as a local real estate story. That's obvious to an editor and a publisher. But when it comes to how do they take advantage of it from a readership and an advertising standpoint, outside help from a company like ours can make a difference.
How do you know what works? Is it merely a matter of driving pageviews?
Not necessarily. Publishers have been chasing views for a number of years, because they didn't really have any other metric to chase, so they're just trying to increase that. And in most cases, that does make sense. But it depends.
Perhaps they don't need additional views. Perhaps they need a better distribution of the content being consumed. Let's say a publisher gets 60 million views a month, but most of them are local and related to subjects around real estate or finance. Our terms won't be ad-related but it speaks to the same things, so we call them "persistent editorial instructions," which acts like DoubleClick for advertising. Instead, we try suggesting placement and timing for certain kinds of content to position them to be picked up further than it otherwise would have been.
Is there any kind of score or methodology you use to determine success beyond pageviews?
If you want to provide anybody a recommendation, or a set of recommendations, the first thing they'll ask to do is sort them. "Tell me what is the most valuable thing I can do next." If you want the most valuable thing you can do next, I need a score as you suggest. We provide that score in three essential ways. For some of the most bold ones, which is not as few as you would think, we simply provide it in monetary terms. No need for me to hide it and anything else. We say, if you run this content here, it will help you drive an additional – just as an example – $172 dollars. Then that's what it is. Monetary terms is the ultimate metric.
Still, some publishers need a metric which they dealt with yesterday and which they believe they'll deal with tomorrow. We'll provide that score in views. "Hey, if you tone down that set of stories in the Middle East in those two positions on the homepage, you stand to generate an additional 248,000 views."
Let's take a good example here, such as The Economist. It might actually not have a direct monetary objective related to views. Maybe their goal is to position their content to drive the most amount of offline subscriptions. That's a more sophisticated goal and we can help folks like that figure out exactly what to program to achieve that objective.
Is there any direct ad sales function that you can do for publishers?
Let's take another client, The Atlantic. You look at their homepage. You find a story that you like. You click into it. You go back. Find another story. You click into it. In that story there's an in link, you click into it, then you go on to a third story. Those five pageviews, could generate 15 ad impressions. There's a different value to the stories that were clicked on. We can connect the monetary values of story recommendations to their ad systems or outside to DoubleClick. Our focus isn’t on getting access to the revenue part of the views themselves. It's this ability for us to inject changes to the environment in mid-air.
How do you get paid? What's Visual Revenue’s business model?
We're selling software to these guys. It's not consulting services.
Here’s how it works: What will happen this morning, 30 people from the New York Daily News, will go to our platform. They'll log in to our terminal. There'll be a set of recommendations for what they should do with their content and some performance reporting. At the end of last month, we launched our social recommendations. It's just another channel that is available to them. We'll be starting with making these recommendations to both Twitter and Facebook.