Media's Valuation Quagmire
Buzzfeed looks to change the narrative by going hard on AI. Media valuations confound the best robot minds. A time for paranoia. Decisions, decisions, decisions.
Welcome to People vs Algorithms #56.
I look for patterns in media, business and culture. My POV is informed by 30 years of leadership in media and advertising businesses, most recently as global President of Hearst Magazines, one of the largest publishers in the world.
In a sign of just how truly mental the media world is at the moment, Buzzfeed's stock more than tripled on news that the company was going hard on AI to “enhance and personalize” content. In a note to staff, CEO Jonah Peretti touched all the media things — iconic brands, tech platforms, cultural impact, award-winning journalism, celebrity, ruthless efficiency — and leaned into the two big things of the moment, AI and “creators.” For a media brand that rose to prominence on a tech meets media proposition custom-built for a new era of socially-driven "content virality," AI mania presents the perfect moment to pounce. The relationship between media and technological change is Buzzfeed’s sweet spot. Go time.
Importantly, they would not put AI work in the obvious way, like CNET/Red Ventures, who were first out of the gate last week, but publicly embarrassed for releasing error-riddled, AI-created personal finance articles. Instead of leaning on this new tech to generate cheap, Google-stuffing fodder, Peretti insisted it would be used to "benefit both content creators and audiences, inspiring new ideas and inviting audience members to co-create personalized content.” AI would underpin better tools for editorial teams and novel new text-based entertainment constructs like a quiz that spits out a romcom about you in less than 30 seconds. Neato.
My take… the whole thing shows how digital media is stuck in an existential quagmire between a fading promise of an old model and a wildly speculative rendering of the future. Today, the only clarity can be found in obeying the basic laws of media land: consumers love video, data is gold, and advertisers always value performance. Profit is the necessary oxygen. Most digital publishers are struggling to match a profitable path to the audience with reliable brand monetization. As a consequence, valuation multiples have nothing to feed off of other than modest turns on cash in the here and now. It’s tough out there.
I admire media innovators trying to figure all of this stuff out especially when they are stuck in a suffocating public company spiral. But a few things stand out to me this sunny NYC Saturday morn. First, is how far we've fallen as an industry when a beleaguered digital media stock can rocket up on news of its embrace of the thing that most threatens its existence — a robot brain that undermines the value of low-end content creation and challenges its primary content distribution point (Google) with the threat of AI clogged pipes and link-killing narrative responses to human questions. But, if you are Buzzfeed, what else are you gonna do? Becoming an AI meme stock seems like a decent option.
Indeed, an industry so backed into the corner by creator-fueled platform dominance and low-yield programmatic advertising has no refuge but to play the AI thing for all it's worth. Sadly, the path forward for scaled digital media business is not as easy as steady audience-serving excellence, brand building, and business savvy. There’s got to be a new trick. Buzzfeed discovered this opportunity once before when the social media explosion and the rise of Facebook’s Newsfeed revealed a potent new surface area for content distribution.
As always, content needed to match the distribution platform. They were the first to recognize that making light, shareable stuff for a hungry social media user could lead to explosive growth. It wasn't a media brand built on an audience passion point, point-of-view, topical or journalistic excellence. It struck gold making shit for social-driven self-expression. Ok, it did news well for a bit, before it became prohibitively expensive to do so. All admirable, BTW. Jonah is super smart and navigates the intersection of content, technology, and modern distribution systems better than most.
And to Jonah’s point as to how best to use AI in media… the idea of leveraging a large language model like Open.AI to make content for a media brand is a tad counterintuitive. Why do I need you to ask the smart chatbot a question for me, just so you can feed it to Google, and fill it with ads and affiliate links? So I can stumble on it later after asking Google about the pros and cons of a variable rate home equity loan? I will do it myself, thank you. The promise of AI is it can answer absolutely any question you have in a silicon heartbeat. It's not having a "journalist" ask the question for you, proofread the results and publish it so you can find it later. One has to add more value.
It seems the Buzzfeed team recognizes AI’s potential as a tool for creativity, and the development of new media experiences, not just a hack that will make the old way faster and cheaper. Of course, the problem with all of this is lots of people can do it. The short-term excitement is predictable. Finding a long-term differentiated position using AI in support of a media model is much harder.
Which brings us back to the digital media conundrum broadly and the other part of this week's Buzzfeed excitement. In publishing, when you can't build a sustainable media business off of a core advertising or subscription-based business model, there is really only one refuge — to become a services company. What does this mean? It means that you are forced to do more and more bespoke things to make money from an advertiser or partner unrelated or tangential to the pure impression value of the media you create. This is the “media brand studio.” This is delivering your specific content creation expertise, your ability to put on an event, to rope in a celebrity or influencer, to do it fast and cheap or against a fading media brand, with or without the core proposition of the underlying value proposition of the media, in the service of building someone else's brand and content need (Chanel, P&G, Coke), probably on someone else's platform (TikTok, Insta or even Coachella). There are good profitable ways to do this FWIW, if you do it well.
Apropos to the service evolution noted above, was the other good news packaged in this week’s Buzzfeed announcement. They would be paid by Meta to recruit and train "creators" to make content for their platforms, an announcement that seemingly brings the company full circle to its social media content-supplier roots. This time Buzzfeed branded listicles and quizzes won't do. They will have to bring real, video-making, influencer humans to the table, train them like an outsourced staffing company and prostrate them to Meta’s content-hungry algorithm. Buzzfeed becomes a human capital provider to a platform. This strikes me as a sorry state of affairs for the industry.
To be clear, the services hustle is not unique to Buzzfeed. We see this everywhere in the media industry, at all the big publishers, including Vox, Condé, Hearst, NYT, Bustle. Media companies becoming media/agency hybrids. Low-yield ad revenue is augmented with high-yield service-driven advertiser offerings. They offer something agencies cannot — the added leverage of a recognized media brand, low-cost content creation, social distribution expertise, talent relationships, etc.
But, here's the catch. While a nimble and highly efficient media/agency hybrid is a perfectly legitimate business to be in, particularly when packed in deep vertical expertise (sport, fashion, tech, health, etc.), it does lead to a conundrum that is playing out all over the markets and in board discussions which I have witnessed first hand.
There was a time, not long ago, when emergent digital media businesses were seen as an important hedge against media model change. This was reflected in the high multiples they were valued at, usually a multiple of the top-line versus the more sober and Buffet-esque bottom. Top-line multiple reflect the promise of the future, of strategic and synergistic potential. Looking back five years or so as a quick reminder: After a foiled attempt to buy the venerable FT, which Nikkei nabbed at 2x revenue and over 30x profit, German media company, Axel Springer, laid down $440M, a more than 13x revenue multiple to acquire the tabloidy digital business news upstart, Business Insider. Vice raised $450M at a $5.7B valuation, also somewhere in the 10x revenue multiple range. Both Buzzfeed and Vox raised money from NBC at around 5x revenue multiples, both without profit.
But service companies typically trade under 1x revenue and well under 10x EBITDA multiples.
A valuation multiple reality check amidst more challenging economic times is no surprise. The bigger issue facing all media companies, particularly digital media ones, is healthy multiples require a clearer view of the future. It's extremely difficult to put a price on something when you can't quite see its money-making role in some future media ecosystem. That would require a clear view of how the content is distributed, of why it is unique and valued by audiences and advertisers. Today none of this is clear, making the cherished revenue multiple incomprehensible and, in turn, bringing huge scrutiny and skepticism to the future stream of profits that these companies promise to make.
It is also no surprise that Vice is the poster child for this pickle. When you have no real profit, a waning revenue growth engine, questionable breakup value, and a business that demands more than a day's work to bring everything in line, you have sorrow. No one is going to pay a premium for a "generational" brand with a slacker’s prospects. The speculated valuation of “under a billion” is optimistic, IMO.
Which leaves us with what's next.
Digital media is morphing at an alarming rate. Some will find a future path as the parental glue that holds together the creator / platform / advertiser love triangle with useful brand service offerings. But this is surely a consolation prize.
Maybe, just maybe, a strong tech/content hybrid like Buzzfeed will find something sustainable and differentiated that adds a more interesting consumer front end to AI machines. Highly speculative.
Or, as you've heard me say, before…. leaner, meaner, definitely smaller, more human, more niche, and all that bla, bla, bla.
Those that are sitting comfortably on top of scaled, highly profitable, diversified, performance / affiliate-driven media business today should do so with a healthy dose of paranoia and reflect deeply on how AI threatens their future. The connection between what a consumer wants and an economic transaction will remain fertile territory for a long while as far as I can tell. Today it means adding layers of content behind a search query. In the future, it may mean owning custom-trained AI models which help inform a consumer about a need and consumption decision better than someone else.
In the interim, promises of a bright future in the form of revenue multiples will be regarded as "check is in the mail" nonsense. Money people will demand a clear and low-risk line to profit now and in the future to generate decent valuations. Unless, of course, you tap the uniformed imagination of bedroom meme stock traders looking for the next Gamestop wrapped in AI.
We made a podcast about decisions and how to make them. It's a banging episode, dare I say. Listen here.
Making decisions in media right now is difficult for all the reasons I waxed on about above and more. Core to any decision is understanding the tradeoffs within. I wrote about this a while back in the context of Instagram’s TikTok conundrum. I also wrote about understanding the Majors and Minors in the media business. Both good counterparts to the podcast.
From the tradeoff note:
1. The larger and more complex the existing business, the harder it is to make trade-offs that risk undermining your core business. This is the territory of classic disruption — introduce a product or service that is hard for an incumbent to imitate because doing so will compromise the legacy franchise.
2. Trade-offs are more difficult to execute when the user base is heavily invested in the existing solution, as is the case above.
At the heart of most trade-offs in digital media are complexities of monetization — balancing the needs of users and the needs of advertisers. In hyper-vertical media they align better. Most of the time you have to find a balance. The web typically forces sub-optimal decision making between short term monetization and quality of experience / long-term loyalty.
Revenue is immediate. The user is invisible. Most web sites trade on a search-to-site distribution model that advances a 'hit and run" mentality. In short, it is easy to trade revenue for experience. That why most web media experiences suck. Further, once you get a taste of that nasty revenue, say floating video players or scammy recommendation links, it's hard to give back.
3. Immediate cash is trade-off crack, once it touches your lips it's hard to put down. Center your decision on what you want to be long term. Pure audience arb is fine if that's your game, but it never lasts. Good media trade-off decisions require vision and patience.
Of course trade-off math is much easier when you have a singular true north. Optimizing a subscription media business demands a singular focus on the needs of the buyer. Trade-offs that advantage acquisition, reduce churn and increase satisfaction are pretty easy to make and measure.
4. Trade-offs are much more difficult in multivariate scenarios. Like the advertising / consumer example above. Things are much easier when you have one constituency to please.
Let's call the next one the Swiss Army fallacy. Generally speaking, the utility of a thing decreases as you add more non-symbiotic functionality. If someone says "You are amazing... you're a total Swiss Army Knife," take it as a compliment and thank the person. They probably mean you are a utility player with a brain and you are very useful in diverse and unpredictable situations. Being a Swiss Army Knife if you are, say, a chess champion, probably means you are not a chess champion. Bad example. "Sporks" have not caught on because they are bad spoons and bad forks. Swiss Army Knives suck. They are bad knives and bad bottle openers and bad pliers. They save space. Ok, I'll give them that. Hopefully you get the point. Trade-offs are precisely that, give something up to make something else more pure, more powerful or more valuable.
5.If you really understand the value you offer a customer or audience, you can make good trade-offs. "Lets do both" is probably an abdication of your responsibility. You may never be great at anything.
There's a difference between a thing that is a platform for many things and a thing designed for a specific purpose. An iPhone is an example of the former, it needs to be a vessel for many applications. In China, the most popular apps are essentially virtual desktops. These are referred to as "super apps", WeChat the best example. Their foundation is payments and communication. Many other applications like delivery services, shopping, and games are built on top of these core functional services. Neither of these should compromise our rule. Instagram is still an application, not a platform. Its virtue is, or was, mostly singular.
But there's a phenomenon that is worth noting here. The "application economy" unlike the web is filled with friction. Building an app users base requires two steps, one to download the app, a second to start using it. This changes our trade-off math. Adding functionality to an existing application may be a more desirable trade-off than embarking on building a second, more focused app on account of the effort to build a new application base. Had the web been architected with native accommodations for identity, payments and advanced device control, things might be different and the movement between applications might have been as easy as a hyperlink. Sadly this is not the case. Adam might like to build a separate photo app and TikTok competitor, but app ecosystem logic discourages it.
6.Your trade-off logic is only as good at the behavioral mechanics that inform it. You are but a pebble in the sea.
I stumbled upon a book on trade-offs. You don't need to read it because it’s one of those business books that makes a simple point in 200 pages. In“Trade-off: What Some Things Catch On, and Others Don’t,” the author argues that business must optimize to either “fidelity” (quality, exclusivity) or “convenience” (accessibility, convenience, affordability). The argument is, focus on one, when you try to do both you will lose to a company who is better at one or the other. Mass-lux is a fallacy. It’s a little like my Swiss Army rule. Or the old product management fav, ”good, fast, cheap… pick two.”
Sadly the world is more complicated than this. Trade-offs are, by their nature, complex decisions that require giving something up to get something in exchange. The only way to get good at it is to ruthlessly question what you are optimizing to. Unfortunately the status quo is not an option for Adam. There are forces at play bigger than Instagram, ones that strike to the heart of why people loved the app in the first place. Finding the next equilibrium will take more vision than plagiarizing competitors, feature by feature.
Have a great weekend.../ Troy
On the topic of decisions: