We’d all like to know what comes next. That can be a spiritual quest, a political one, or in the case of news publishers, one that would help them know what it is readers who land on their site would like to read next.
A new batch of news-oriented tech companies are hitting the marketplace, claiming to better understand — and help news publishers act on — what readers are more likely to read next. Know that, and publishers can better satisfy those readers, getting them to click on more pages, providing more ad-targetable data, and growing their businesses on the relative cheap.
It’s another riff on publishers’ renewed concentration on core readers, satisfying them more deeply and getting them to spend more time on site and maybe even pay for a digital subscription. It’s another way of saying we need more revenue per customer (The newsonomics of ARPU). New customers aren’t being minted overnight (most news sites’ unique visitor growth has stalled), so it’s time to improve the experience of current customers.
In macro terms, if the U.S. newspaper industry could use these newer technologies to improve revenue by 10 percent, that would be worth about $300 million a year, collectively. That’s a big number in the new no- to low-growth era we’re in.
Three companies — YieldBot, JumpTime, and Outbrain — are among those who offer news and media companies differing approaches to better reader engagement. Each has its own story to tell, and each is emblematic of a wider trend in the industry: mastering how the digital business is oh-so-different than print.
They have a common belief in the limitation of human intelligence, especially that of the common editor.
The human reader brain, goes the theory, is far more unpredictable than editors ever believed. (And most editors I’ve known have a low, low regard for readers’ brains.) Online, editors and designers have long packaged stories, placing whatever stories seemed relevant to the story being read, either manually or by simple algorithm. Of course, that’s worked — to some degree.
To how great a degree now is increasingly measurable. And these would-be tech partners are telling publishers: you are leaving stories (and money) on the table. We’ll show the real, provable relationships between stories and by adjusting your presentation, you’ll bump up your pageviews — especially among valuable core readers — and make more money.
It’s an evolution of thinking beyond a single edition of a newspaper, read from front page to back, in an orderly fashion. Few people actually read papers in that orderly a fashion, but that’s how publishers and editors thought about it. Then they constructed their online sites the same way, with disproportionate attention to the hallowed home page, some attention to the section “fronts” and less to the “article” pages.
These companies are trying to build divining rods for the digital age.
Now as sideways traffic — greatly multiplied by Google, Facebook, and Twitter links, mentions and touts — has become recognized as the way things really are, publishers need new understanding. There are real, discernible patterns of behavior, if you crunch a lot of data, and these companies can show it to you on graphs, scatter charts, clusters and more. Let your online presentation people link stories or sections they wouldn’t have otherwise linked. Let your audience management people make longer-term decisions about how valuable that Facebook traffic was compared to that Google traffic. Let your ad staff have the ammo it needs to prove out the kind of visitors who are attracted to certain site sections.
“Sure, we’ll show you where your readers come from and where they go off to, but the real question is what they do when they get to your site. What’s happening at this moment,” says Jonathan Mendez, CEO of New York-based start-up YieldBot. Mendez, a self-described “crusader for relevance,” was a principal at ad optimizer Offermatica before it was sold to Omniture in 2007.
L.A.-based JumpTime paints itself as an “optimizer.” It, too, is focused on what happens once someone hits a news site. JumpTime is all about better routing. Using its “Flo-Power” metrics, the company assigns value to different pages not just on the basis of its individual usage, but largely on how well that page performs in redirecting traffic elsewhere on the site. Some pages are freeway cloverleafs, others not so much. “Why would you send someone to a cul de sac?” asks Michele DiLorenzo, JumpTime’s CEO, an MTV business development veteran.
Both JumpTime and YieldBot give you real-time info on which pages are creating high bounce (exit to other sites) rates, for instance. “A page,” says DiLorenzo, “is a determinant of what happens next.” And wouldn’t we all like to know that?
Outbrain takes a different approach to the same issue. It produces seemingly simple modules of what appear to be “content recommendation” links. Aggregate Knowledge, Inform Technologies, and Sphere (first bought by AOL, rebranded and then sold to Outbrain this year), among others, have plowed similar territory, but never got much traction with news publishers. Outbrain says it produces links that are “things of interest, not just related.” It’s another complexity-reducer, applying its algorithms to individual and group reading behavior, maybe offering a Spanish debt story or a how-exercise-will-save-your-life article to someone reading a Jihadists-target-Letterman piece. Counter-intuitive, but customers say it works to juice traffic.
Outbrain is a five-year-old New York-based company, which just launched its mobile product. It enables publishers to optimize site usage, like YieldBot and JumpTime, and also offers several networking features, recirculating both traffic and small revenues among its numerous member publishers. It has the most established base of customers.
Both JumpTime and YieldBot are newer entries, now testing their products in beta with a few major publishers, looking for an edge. Each site offers consoles and tools to manage data, draw intelligence from it and make real-time or over-time decisions from it. JumpTime works on a fee-for-service model, while YieldBot and Outbrain focus on revenue shares.
The mantra of these companies: Counting traffic with Omniture, Chartbeat, and Google Analytics is oh-so-first-generation. They will tell you how people got to your site and where they are going, but it’s tougher to get real, actionable intelligence about what they are doing on your site, and what they might do, given different content choices.
“Publishers don’t know the [relative] value of their assets,” says JumpTime co-founder Anke Audenaert. “Can you imagine operating a business and not understanding the value of your assets?”
Of course, publishers thought they knew that. They thought, though, of value created by whole editions, circulation value and ad value. Now as readers flit from one story/video/blog on one site to another on another site, all of these tech companies are saying, essentially, value has been atomized. Each article or page has a value, and that value grows or diminishes in context to other articles and pages. And you need our algorithm to figure it all out.
Of course, any good algorithm produces a better yield. Outbrain will tell you its algorithm provides a 6-to-8 percent lift in pageviews for sites that properly deploy its boxes of linked stories; some of its customers tell me they’ve gotten that and more. JumpTime will tell you it can improve revenue by 20 percent by maximizing that secret sauce of FloPower, creating that much more engagement.
These companies, and others like them, are in the business of modern divining rods. Divining rods for the digital age.
Mendez says it is all about intent, determining a web users’ intent better — and then satisfying it. “People have a goal-oriented state of mind,” he says Mendez. “Something is motivating them. That intent makes the medium good at demand capture.” He says big publishers, especially, have enough data to help advertisers better target readers; they just need to use it much better. “Media is worth much more than they are getting.”