(First in a series. Post two, on Twitter’s solution, is here).
Marketers – especially brand marketers: Too many of you have lost the script regarding the critical role you play in society. And while well-intentioned TV spots about “getting through this together” are nice, they aren’t a structural solution. It’s time to rethink the relationship between marketers, media companies (not “content creators,” ick), and the audience.
So let’s talk about it. Grab your favorite beverage and read along. I’m heading into a bit of media theory for the next couple thousand words – I hope this will start an interesting conversation.
For those of you who want a TL:DR summary, here it is: It’s time to get back to the work marketers used to be really good at: Deciding on the appropriate context in which to engage your audience. And it’s time to pull back from a habit most of you have fallen into: Letting the machines choose your audience for you. Thanks to new approaches which fuse at-scale ad targeting with high-quality editorial product, you can step into this renewed role without sacrificing the reach, precision, and targeting afforded by the likes of Facebook, Google, Twitter, and their kin. To understand how, let’s review some history.
The Old Media Model
If you read this site back when I wrote regularly on media (roughly 2003-2015), you’ll recall I laid out several basic tenets about how the media business works. It’s comprised of three core components: Editorial (the media company’s content), Audience (people who give their attention to the content), and Marketer (commercial actors who desire the Audience’s attention in the context of the Editorial). Of course, in the past ten years, a fourth component has eclipsed all three: The Internet Platform.
Before the major Internet platforms deconstructed the media business, the three original components came together in what we’ll call a media product (I’m still partial to “publication,” but many think only of print when they hear that word). Print, television shows, and early web sites all served as vessels for a commercial relationship between Editorial, Audience, and Marketer. The media company took the financial risk of creating and distributing the media product, and if successful, the marketer paid to run advertising inside the media product. In some cases, the audience also paid a subscription fee for the editorial. But for most media companies, advertising support was crucial to chin the bar of profitability and make a go of it as a business.
A critical element of the media-product-as-vessel model for commercial transactions was that context matters. The media product created context for audience engagement, and if the marketer offered messaging that aligned with that context, it stood to reason that the audience would be more receptive to the advertiser’s message. Suffice to say that with the rise of audience buying on massive platforms, context has been lost, with nearly incalculable downsides across the media ecosystem (and society at large). More on that later on.
Meanwhile, back in those pre-platform days, distribution was important, but it was also a constant. Most media companies consolidated distribution by acquiring broadcast licenses or cable networks (for television) or print distribution networks (if you were a magazine or newspaper company). And if you were a media startup, you could leverage those distribution networks for a relatively predictable rent – often without spending any capital up front. When we started Wired, for example, we secured newsstand distribution by agreeing to split the revenue earned by our nascent magazine with our distribution agent.
I call this old-school model “Packaged Goods Media.” Fifteen years ago I noted that “PGM” was giving way to a new model, which I termed “Conversational Media,” or CM. CM, of course, was the precursor to “social media” – Twitter, Facebook, YouTube – and as I thought out loud about this new phenomenon, I noted several crucial distinctions between it and Packaged Goods Media. I predicted that the economics of Editorial, Audience, and Marketing were all going to change dramatically. In many ways I was spot on. But in several others, I was dead wrong. Here’s a summary of a few key points:
- Editorial models would evolve from “dictation” to “conversational,” where the audience – and knowledge of the audience through data – became a central driver of editorial creation.
- Distribution would become nearly free, obviating the rent-seeking monopolies held by major media companies. In fact, I wrote: “economic differentiation based on the control of distribution – the very heart of PGM-based business models – is irrelevant in CM-based services.”
- Online, publications become more like a service, rather than a product. I noted that software, which was still largely a packaged product, was also heading in this direction. That means media will have different economics and different advertising models over time (I called them “native advertising” at the time).
I’d argue that over the next ten years I got the first and third predictions relatively right, but I entirely whiffed on how distribution would play out. I simply failed to imagine how Facebook, Google, and others would leverage their newfound control of audience attention. In one piece from 2006, I wrote:
“…finding massively scaled Conversational Media companies [besides Google] is a rather difficult search … it’s unclear whether CM companies will mature into massive conglomerates like Time Warner.”
Well, it’s certainly clear now. Facebook, Google, and their peers are among the most powerful and well-capitalized companies in the world, and they got that way by doing one thing very well: Capturing the attention of billions of us. That gives them a near monopoly on digital distribution, which they’ve leveraged into a near monopoly on digital advertising. In the process, these tech platforms have eliminated the traditional role of publishers as a proxy for audience interest and engagement. I used to believe this trend spelled the end of high-quality independent media brands – indeed, it’s why I didn’t start a media brand after selling Federated back in 2013. But media models are always evolving, and I now see a new way forward. To understand that, we must first review where we stand today. And to do that, we must examine arbitrage.
If I were writing a sequel to “The Search” focused solely on how digital media models have shifted in the past 15 years, I’d probably title it “The Arb.”
It would not be a pretty story. In the past ten years, audience arbitrage has become a dominant model of the digital media business. It’s an awful business practice that erodes trust, devalues media brands, and dilutes the importance of marketing. What follows is a bit of a rant, but hell, you’re still reading at this point, so refill your glass, and let’s get to it.
The dictionary definition of arbitrage is “the simultaneous buying and selling of securities, currency, or commodities in different markets or in derivative forms in order to take advantage of differing prices for the same asset.”
In media, the asset being arbitraged is audience attention. The arbitrageurs are publishers. Their enablers are the major tech platforms, fueled by dollars from advertisers.
Here’s how it works. A big publisher like Buzzfeed or Cheddar sells a million-dollar advertising deal to a marketing brand. The media company guarantees the marketer’s message will collect a certain number of audience impressions or views, charging the marketer a “cost per thousand” for those impressions. (Known as “CPM,” cost per thousand pricing ranges widely, from a few pennies to $25-40 for “premium” placements). Utilizing a Packaged Goods media model, the publisher might fulfill those impressions on its “owned and operated” properties, but over the past ten years, doing so has accrued significant drawbacks. The top three:
- It’s expensive. Acquiring and retaining audiences on a media company’s own property is often far more costly than finding those same audiences on an at-scale platform like Facebook or Google.
- It lacks sophisticated targeting. In the past decade, marketers have grown accustomed to the data-rich precision of large platforms. They don’t want to pay for just any old Buzzfeed or Cheddar audience member. They want their messaging to reach exactly the target they specify, and most publishers don’t have either the technology or the audience scale to fulfill the data-driven demands of modern marketers.
- It forces extra work on the marketer. I am not the first, nor will I be the last to note that marketers and agencies don’t like to do extra work. While plenty of larger publishers have built high-quality advertising solutions on their owned and operated channels, marketers view these point solutions as just one more channel they have to manage, analyze, and report on. It’s just So Much Easier to buy Facebook, after all.
Because of all this and more, publishers have become audience buyers on Facebook, Google, and other networks. Enterprising publishers began packaging their own content with marketing messages from their sponsors, then they got busy promoting that bundle to audiences on Twitter, Facebook, and Youtube, among others.
This is where “the arb” comes in: The publisher will charge the marketer, say, a $15 CPM, but acquire their audiences on Facebook for $7, clearing an $8 profit on every thousand impressions.
You might ask why the platforms or the marketers don’t put a stop to this practice, and you’d be right to ask. But consider the economic incentives, and things get a bit more clear. The platforms are getting paid for what they do all day long: the delivery of precise audience impressions at scale. As far as platforms are concerned, the media brands are just advertisers in different dress. Over the years, Facebook and Google have even accommodated the arbitrage by connecting all parties directly through their advertising technology systems.
OK, so the platforms get paid to deliver audiences to marketers on behalf of media companies, but why on earth do the marketers put up with being arb’d? Couldn’t they just pay the same $7 CPM directly to Facebook, eliminate the middle man, and save the $8 spread?
Well, indeed they can, and in most cases when it comes to buying audience on Facebook or Google, that’s exactly what they do. But remember my comments about context way up toward the top of this article? Some marketers still believe that the context of a media brand can help their messaging perform better, and they’re not wrong in that belief. So they’ll pay a bit more to have their messaging associated with what they believe is quality editorial. And if that media brand does the work of acquiring that audience for them, so much the better – that’s less work for the marketer to do.
But let me be clear: arbitrage sucks. Arbitrage is only lucrative in markets with imperfect information. It’s usually a great strategy in the early stages of a new ecosystem, when media buyers are less familiar with how advertising technology works. As those buyers get smarter, they start to squeeze the media company’s margins, devaluing content and context, and pressing ever closer to the price they could get directly from the platform. A good example is Demand Media – a company that, a decade ago, managed to insert itself between Google’s search algorithms and an advertiser’s desire to be associated with content around a particular topic. Demand pulled off a billion-dollar IPO based on creating advertiser-friendly “content farms” around popular Google searches. But advertisers figured out the arb, and Demand’s once billion-dollar valuation fell more than twenty fold in the past five years. A similar fate has befallen the once high-flying arbitrageurs of social media. Cheddar, Vice, BuzzFeed, and many others all played the game, but over time, markets will root out an arb. (Cheddar was smart enough to sell before its arb was uncovered – but it sold at a fraction of the sky-high valuations its peers once held).
But wait, one might ask – aren’t the media companies adding true value? What about that context, which makes a marketer’s message more relevant and engaging? Isn’t that worth something?
It certainly is, but this is where the lack of transparency around ad buying on platforms comes into play. Audience buying is cloaked in opacity – the major platforms are deeply invested in making sure no one truly understands how attention is priced. That means a media company buying audience on Facebook or Google will always be at an informational disadvantage – exposing them to a new kind of arbitrage, one executed by the platform’s own algorithms and benefiting the platform’s bottom line. Again, arbitrage works best in markets with asymmetric information features – and informational asymmetry is built into how Platforms operate. Over the past five or so years, most major media companies have come to realize they’re the ones being gamed.
Audience arbitrage on platforms has even more destructive attributes. Because media buyers have outsourced their audience acquisition to either the media company or the platform itself, the marketer becomes disconnected from the context of its audience. Millions of impressions are scattered across millions of tiny content bundles, all of which are lost in a sea of endless posts on nearly every imaginable topic. The context and meaning that holds all brands together is lost. Media companies, pressed by ever-thinning margins, will cut corners, buying “junk traffic” or worse, creating junk content that titillates or tricks audiences into false engagement. On the surface, boxes get checked, audiences get served, impressions get logged. But over time, editorial content deteriorates, deep relationships between brands and audiences attenuate, and the media ecosystem begins to fail.
So what can be done about it?
Well, at The Recount we’re exploring a way forward, through a brand new partnership we’re launching on Twitter this month. We’re calling it “Real-Time Recount,” and in the next installment of this post (I’m pushing 2500 words here, after all), I’ll explain more about the theory of the case behind it. For now, you can read more about what we’re doing in this Ad Age piece (paywalled, alas), or over on Fred’s blog. Thanks for coming along, and I look forward to the conversation I hope this will spark.