Don’t just survive: A three-step publisher strategy to beat Facebook at its own game
by Matt Crenshaw, Global Head of Distribution, Outbrain
Get a group of digital publishers together, and you won’t get five minutes into the conversation without someone addressing the elephant in the room — Facebook. You’ll hear that their audiences are spending more time on Facebook, advertising dollars are getting vacuumed up, and that Facebook is unilaterally resetting the terms of their relationship, sending dwindling audience numbers to publisher sites while politely threatening to hold their content hostage with Instant Articles.
But that’s not the surprising part. The surprising part is what comes next, when publishers say: “how can we work more with Facebook?”
Publishers publicly treat Facebook like their oldest frenemy, but in their own board rooms their smartest minds can’t map out an endgame that leads to a positive conclusion.
Great Content is Imperative — But It’s Not Enough
Let’s recognize that publishers have to capture the hearts, minds, and attention of their users to survive. They do that everyday with content, and there’s no substitute for creating great content, any more than a restaurant can find a substitute for serving great food.
But just creating more content, or better content, isn’t enough.
While everyday people should recognize the shining light of true, civic journalism and reward it, that’s not the reality. If digital publishers want to win, then media companies need to think bigger. In fact, they need to think like tech companies.
Publishers say they want to be better at data than technology companies. That’s the right thing to say, but understanding the call-to-action that lurks beneath those simple slogans is what’s missing today. To do that, we should focus on the three details that Facebook itself thinks matter most – Personalization, Data and Automation:
The magic of Facebook is not simply social, but lies in its personalization. Every person who logs into Facebook sees something different. The same goes for social platforms like LinkedIn, Instagram, and Twitter. We take this customization for granted, but it’s personalization that makes Facebook irresistible.
Now think about what happens when you log into the New York Times. Even if you’re a paying subscriber and registered user (which, I urge you to be), you see the same news that anyone else sees. Granted, news is a different animal from cat photos and baby updates: publishers have a journalistic responsibility to deliver the news of the day. To achieve any semblance of an informed citizenry, we need a common view on what’s important.
Like most publishers, the deep levels of personalization are absent in the actual content, and appear only in the advertising. For example, the Times relies on deep layers of user targeting, creative media testing, and personalization, even though most of those ads are completely ignored.
Despite generations of attempts by media companies, personalization remains largely uncracked in publishing. As with Facebook, users don’t want a generic version and a personalized version. They want one version. The one they see every time they log in, that adapts to the mode they’re in, the device they’re on, and delivers the content that slots seamlessly into whatever narrative path they as a user are on — whether they’re an expert in international relations or reading names like Trump and Putin for the first time.
2) Data – know the value of content
Digital publishing may look as easy as creating content and waiting for the ad dollars to roll in. But the reality is that digital media is a massively inefficient business to run. There is so much inefficiency to be squeezed out of the current ad model, that it’s almost impossible to blame the business model alone.
Every dollar a publisher makes ties back to content discovery. A user thinks “Hey, this link looks pretty interesting, I’m going to click it…” and whammo!, they go to the next page. When the next page loads, so too do the ad boxes, the native advertising, and all of the calls-to-action to download or subscribe. Call these the “revenue events.” A publisher makes ALL of their revenue when a user calls a page–by clicking– and these revenue events are dependent on the user, their demographics, geo location, content type, ad key value, targeting, etc etc…”
So, the game of digital publishing now looks slightly more complex: it requires matching the right user to the right content, where that content carries the most revenue. Those are the three plates publishers are always spinning: users, content, and revenue.
Of course these three plates aren’t managed by a single system. Publisher tech has evolved three totally separate systems with three separate objectives. No shared metrics, and operated by different teams inside the publisher.
• The Content Management System and Web Analytics to manage content.
• The Content Recommendations Engine to match that content to the right user.
• The Ad Server and SSP to deliver revenue when the content is loaded.
This scenario is almost unfathomable to someone who has never worked in publishing. In any other industry, the inability to understand the revenue value of your core product would lead to constant flailing, zig-zagging on strategy, and blindness about how much of the product to create, or any quantifiable way to improve upon it. We are tackling this problem head-on with our recently launched revenue metrics that will help publishers unify the disconnectedness between users, content and all their sources of revenue.
Publishers are getting more and more data, for sure. Talk to any senior-ish level person inside a publishing company, and they have eight tabs open on their browser at any one time with eight different dashboards.
But none of these dashboards answer the most fundamental question: “What is my content worth?”
With so many different types of ads, targeting, and decision engines running on a publisher’s site, the revenue generated from a piece of content has become totally opaque, buried beneath various auctions and impossible to retrieve. So much so, in fact, that most publishers have stopped asking this question, settling merely for what’s available rather than what’s required. Our new revenue reporting will bring these disparate views together.
However, what’s currently available to most publishers are lots and lots of averages. Publishers can see look-back averages for RPMs across a section, a site, or their entire network, sliced by day or by week. This fills up lots of spreadsheets and provides decent diagnostics. But a data-driven business cannot be run on averages. That would be like trading stocks based on the average over the past week, while someone else is trading based on precise real-time values–an unfair fight.
Having this accurate, granular data is the alley oop; the automation to act on that data is the slam dunk.
Executives and team leaders need the visibility and data transparency that dashboards provide. But if you’re looking at dashboards because you’re wondering who in your team can do what with which data, then I’m betting against you.
Let’s go back to our three-part puzzle, the Publisher Stack. The Content Management System is great for letting content creators and capture and publish their work, and sometimes, especially with the new wave of homegrown systems, can provide data on what content is trending or would be best to create. That’s a helpful nudge in the right direction, but having great content that no one looks at — as we’ve seen — drives zero revenue. The question has to actually be splintered into a million derivations of that question: what is the best content for each individual user.
Remember, you can’t run a successful data-driven company based on averages.
Now, let’s look at the ad stack. The ad server, header bidding, SSPs and other programmatic partners ecosystem — and there’s a new breed of programmatic superstar inside the publisher org that has started to translate this mess into clear outcomes. The fascination with auctioning off every pixel of the page in real-time to generate the highest yield, that’s the game.
Despite all the innovation in the ad stack around increasing yield on a given page it is still critically stupid when it comes to actually making that page view happen. Today, programmatic is still a lean-back-and-wait game for publishers. Wait to see if a user will go to a page of content. Then, as soon as they click to call the page, all the programmatic bidders come out of the woodwork and in milliseconds compete to be the winning bid composition on that page.
Needless to say, the ad stack and the content stack don’t talk to each other.
If you’re running a digital publishing company, that line should make you wince. But stroll the halls of a publishing company, and you’ll see that the editorial folks don’t sit with the ad ops folks. They don’t use any of the same systems, and rarely are they even held to the same goals. In the kingdom of digital publishing, every internal team stays within its own fiefdom.
But wait. There, in the publisher stack, between the content stack and the ad stack, is something in the middle: the content recommendations engine. Historically, these have been the partners living at the bottom of the page, serving up related links and “You Might Also Like” to users after they finish what they’re reading.
Personalizing the next click for users is an extremely important problem to solve, and has led to more engagement and new revenue streams for publishers (through paid recommendations in the form of advertising). Once the user comes to the publisher, it is the publisher’s job to maximize the session and revenue from the experience (just like Facebook does for their users). What Publishers don’t often consider/know is that doing so relies on data, technology, and automation that understands and factors in the holistic value of every action a user takes on a page.
Some of our publishers see content recommendation engines as a challenge to display ads, since they can create an alternate stream of revenue apart from the display and video advertising complex. The whole native vs. display vs. video argument. Nothing could be further from the truth. ALL of these things must work together. The publisher must know the outcome of any one click (or trade-off to a user leaving the site). Therefore, the real enemy of the content recommendation engine is that digital publishing has been an inefficient business where publishers haven’t had the most basic personalization, data, or automation tools to best serve their audience… and serving audiences is the cornerstone of building big, profitable media businesses.
Since content consumption triggers all monetization on a page, we’ve focused the last 18 months on creating solutions for these exact problems through our new product Automatic Yield.
Incremental updates and short-term thinking in digital publishing isn’t a winning strategy. Publishers need to understand the revenue value of their content, incorporate data into their decisioning and have a single place to measure the financial health of their business. Facebook isn’t the answer, but with the right insights publishers can thrive, and even compete, with social platforms.