Shooting The Messenger (Or how to turn a $50 million investment into $3 million of revenue)
Welcome to my post-Substack newsletter*. The title is a delightful phrase I pilfered from Eugene Thacker's Infinite Resignation and accurately describes the theme here.
This edition is devoted to shameless media navel gazing. I keep being told that the news business is dead by people who point to the atrocity below as evidence. To be fair, none of the people saying this work in the news business, but I think there are lessons even for people who understand why The Messenger was destined to fail, and for reasons that have nothing to do with secular economic developments or the state of the digital ad market. So bear with me:
Shooting The Messenger
(Or how to make $3 million in revenue on a $50 million investment.)
Let's start with some stipulations:
1) Early stage venture capital is the riskiest asset class available to an investor (not including idiotic cryptocurrencies, which I think of less as an asset class than a screening mechanism for people easily parted with their money).
2) Early stage VCs back founders more than ideas or models, and founders who have sold a company before are already "validated" in some respects. Whether prior experience is relevant topically or temporally (meaning, is it relevant now?) is not always a consideration.
3) Digital media plays are often wrongly classified and valued like tech companies, even though they don't scale the way, say, a software company does. This creates pressure to scale at a speed that is often untenable and incentivizes delusional thinking about market size.
4) There are plenty of business models that worked circa 2008 that are simply not viable now because they were always a losing proposition in the long run. One is aiming for high volume traffic and monetizing exclusively via low CPM programmatic ads.
So how is all of this relevant to the Messenger? Let's work backwards from the business model:
The Model
Part of the reason why reach + programmatic doesn't work anymore is because traffic was easy to game in 2008, and less so now, and more importantly--and I cannot stress this enough--traffic and audience are not synonymous. You can build a sustainable media property here in 2024 if you have real audience, and you have a relationship with them and they trust you. You can also monetize the most loyal among them in many different ways. Traffic, however, is not de facto audience. If you land here because you searched my name but were looking for a dermatologist who is also named Elizabeth Spiers, you are traffic for this post, but you're not my audience. If you landed here because you were searching for me, you might be my audience because you actively wanted to read something I wrote. You might come back, subscribe to this newsletter, even opt for the paid option.
The disparity between casual traffic and audience is especially relevant for media properties built on paid media strategies and gaming social platforms. Paid media strategies don't guarantee any kind of real engagement, and they don't guarantee that people are even reading or viewing content in any meaningful way. The threshold for what constitutes viewability--the metric platforms use to determine whether someone actually saw a video or article--is extremely low. Scroll slowly on Facebook or leave your browser open for a minute, and you have "seen" a Facebook ad, whether you really registered it or not. If 50% of a display ad was viewable on your screen for a second, or two seconds if we're talking about video, you "viewed" it. Even click through rates as a measure of real interest are not reliable because any headline writer worth their salt knows how to generate an irresistible hed.
Organic social media is a bit different in that you can use it to build audiences because a really good SM account builds trust and loyalty the way any other kind of media does. But if your model is high-reach low-CPM programmatic, you're not necessarily going to capture that user on your destination site, and almost never on the homepage. If you want to build and sustain audiences you need to be operating in different modalities, and you need to meet them where they live, and your business model needs to include strategies for monetizing those communities when they prefer to engage off the flagship property.
So that's the first problem. The Messenger was a 2008 era business model trying to exist in 2024 by building traffic (via junky aggregation features, clickbait, etc.) but not audience, and monetizing via programmatic display ads.
Why are these people trying to make 2008 media work in 2024? That brings us to the founder problem. Jimmy Finkelstein, the founder, started the Hollywood Reporter and also The Hill, two very different organizations. HR is a trade publication, and a good one. It's easy to monetize because people in the industry have to read it and their employers pay for it. I love trade publications; the business model is straightforward and reliable.
The Messenger is not a trade publication; it's a consumer publication, and Finkelstein had ambitions of building it to compete with some of the largest news orgs in the country. That's where The Hill comes in. At one point in its history, The Hill functioned a bit like a trade because it published insidery beltway scuttlebutt, but Politico and other pubs ate its market share for lunch, and it became more of a consumer play, driven by ... clickbaity articles. It was less ersatz Politico than ersatz Daily Mail. It was able to achieve reach (traffic, not audience) and there was a promising events business on the backend.
The Founders
Finkelstein wanted to replicate that with The Messenger, and brought along Neetzan Zimmerman, the viral content guru who had developed The Hill's content farm/traffic engine. He hired Richard "Mad Dog" Beckman**, who left Conde Nast circa 2010 after years of running ad sales there when he had trouble adapting to changes in the ad market and got demoted to run Fairchild's ad sales.
The two of them together are not exactly the picture of media innovation in 2024. Zimmerman knows far more about the way the Internet works than they do and even what he's doing is outdated. I teach a class on the topic at NYU and the people I bring into the program to talk to my students are working on things like using local LLMs to scrape public documents, building digital products that provide solutions to civic problems, and testing new business models that focus first on audience needs.
But this is the team that got $50 million in funding, because "experience." And let's be honest, they're white guys of a certain age who fit the profile of the white guys of a certain age who funded them. (Insert culture fit bullshit.) The fact that a reach play monetized by programmatic ads is a laughable thing to build now was considered irrelevant because investors assume that prior success is indicative of lower risk. And possibly that white guys of a certain age are lower risk. (I joking refer to this as Demographic De-risking, but as a female founder who's been told I'm too high risk despite having some successes under my belt, it does periodically make me want to move to a desert island with no internet connection and never think about this industry again. But I won't because I'm a masochist.) A bunch investors basically paid a couple of horse and buggy specialists $50 million to build a Ferrari.
Which brings us to that number. FIFTY. MILLION. DOLLARS. Let's say spinning up a reach-based digital media property that monetizes via programmatic is a good idea. (It's not, but humor me.) How much money does it take to do that? Do you need a giant fancy office space (Class A, no doubt), investigative journalists (not cheap!) extravagant T&E budgets (to woo the rare direct sales client), a bespoke custom CMS (kill me). No, you don't. If I were going to do that I'd probably still hire Zimmerman, make everything remote, find Internet junkies who can write and make video but don't have to be trained journalists, and let them all do their thing. How much does that cost? If you just wanted to build a traffic machine? I think you could do it for under a million, plus whatever Zimmerman costs these days. But what would be the point?
So there's a big disconnect here between Finkelstein's stated ambitions (compete with big national outlets), what they produced (mostly aggregation, traffic bait), and how much it cost them to produce it.
Lastly, there are different types of funding for these sort of operations, and venture funding is both the hardest to get and comes with the biggest expectations for returns--in part because VCs often value digital properties in ways that more resemble tech companies. The platonic good media investor is someone who understands the media business (and that it doesn't scale like software), understands the product (and can tell the difference between good content and bad, which is depressingly rare in my experience) and has high enough expectations for return that they have an incentive to push the organization to be ambitious, but are not delusional about what the limits are. They understand that a good return for a media company is not what a good return for an AI company would be. Those people are usually angel investors and former media entrepreneurs themselves, and they're hard to find. Venture money is prestigious but often lends itself to growth expectations that can only be fulfilled by taking shortcuts that are damaging in the long term. Part of audience building is creating a feedback loop where you use inferred and reported data and audience interaction to understand whether what you're doing is successful and what the audience is or isn't willing to pay for. You can't do that with a seven month runway, which is how long it took Finkelstein to run through that $50 million.
I have no shortage of other navel-gaze-y thoughts about the media business and how to build media orgs that work and will no doubt get them out of my system here sooner or later. In the meantime, thanks for reading, and if you'd like to support my work here, there are two options: you can contribute at any rate, or subscribe for the year, in which case you will also receive a quarterly little print zine I'm making. (You can subscribe directly here.)
Also: below are gift links to some of the NYT columns I've written lately. You can also hear me discuss financial news of the week on the Slate Money podcast with my co-hosts Felix Salmon and Emily Peck here.
* I left Substack after Hamish openly promoted Richard Hanania. While any big platform will inevitably be swarmed by aspiring Nazis, the CEO does not have to actively promote them, and that is a choice.
** "Mad Dog" because of his insane temper, not like your cousin Ray who earned the moniker because he once took a ton of ketamine and decided to impersonate a cop to a cop while conspicuously pants-less.
Recent work in The New York Times:
The Miseducation of Nikki Haley (Haley attended a segregation academy and so did I. Here's what we learned about the Civil War there.)
I Figured Out What ChatGPT Reminds Me Of (Why AIs are like small children, and training them seems overly familiar for those of us who are parents.)
The End of Snow (On climate change, existential terror, and the last snowman in Omaha.)
The Reason Why People Like George Santos Lie About Nonsense (On famous liars.)
A Tech Overlord's Horrifying, Silly Vision for Who Should Run the World (On Marc Andreessen's "manifesto", which has the pathos of the Unabomber manifesto but lacks the ideological coherency.)
P.S., I'm also teaching an online opinion column writing workshop on March 6th, from 3pm ET to 6pm ET. Details and registration are here.