Discover more from People vs Algorithms
BuzzFeed's Dirty Laundry
What the shift to Creators and AI portends for the biz. Plus, the digital ad squeeze.
Welcome to People vs Algorithms #61.
I look for patterns in media, business and culture. My POV is informed by 30 years of leadership in media and advertising businesses.
Sometimes it’s nice to read in the browser.
BuzzFeed is less bellwether, more avatar of the many hacks digital media has relied on over the past decade. Now that it’s public, we get to see its — and by extension the publishing industry’s — dirty laundry.
Set aside the economic “headwinds”. Set aside the awkward BuzzFeed / Complex combo that brought together two companies with fundamentally different approaches to media value creation. BuzzFeed is the OG viral media company. Lot’s of good people work there. Its struggles are worth examining to understand the path forward.
Q4 '22 results released last week were sobering. While full year 2022 exceeded ‘21 by 10%, Q4 showed the business losing momentum. Ad revenue declined 27% vs Q4 ‘21. Content revenue (a mixed bag of ad content creation and production revenue) was down 9% YOY. Commerce revenue, mostly affiliate, was a bright spot, up 67% YOY but representing less that 30% of total revenue. Q1 projections look equally bleak, guiding to $61-$67M in revenue, an $18-25M adjusted EBITDA loss (Q1 is always tough in digital media). The revenue estimate represents a 30% decline from Q1 '22. All told, the business burned $22M of cash in ‘22. With only $56M remaining, a market cap of just two times quarterly revenue, something’s gonna give.
BuzzFeed’s founder and CEO Jonah Peretti offered a telling perspective on the call, proclaiming "The future of digital media will be defined by creator led and AI powered content.” When times are tough, sometimes you have to do a look-over-here to find breathing room. But, something in Jonah’s statement feels prescient. And it suggests media companies, especially digital ones, will look a lot different than they do now.
Particularly for publishers that have 1) limited prospects of building a subscription business, 2) an affiliate or marketplace business that is a small percentage of overall revenue. The next move? Peg your future to an ecosystem dominated by creator-led short vertical video (see the TikTok-ization of everything: Insta, YouTube, Snap, Spotify soon Pinterest). Cross your fingers that these platforms find an equitable monetization formula that pays media brands to compete with the rest of humanity.
The decision comes with consequences. Your media brand becomes a thin layer on top of “creators” (think journalist x talent), enough to maintain audience expectations. Your brand continues to play the curatorial role, but evolves to share the spotlight with talent brands. In modern journalism, innovators like Puck are defining brand around visible talent “partners” that do the work. In entertainment, the media brand (NBC, A&E) never meant as much as talent and show brands. In the end, it’s just a twist that reflects how the power balance has shifted from institution to the individual.
Next you have to come to terms with "three-way" economics. You are renting creator talent to feed someone else's platform. Dollars get divvied between you, the talent and the platform. Leverage evaporates. Your media business better be efficient to feed hungry partners.
The new media company focuses on video format development (tune into podcast next week as we discuss this), talent management, live activations and distribution platform relationships. Monetization, still important, becomes more of the platform's role. You employ fewer sellers and marketing types and get much leaner.
If you are like Jonah, and think that AI is a potential savior, you focus your tech efforts here, away from managing complex content management systems designed to ferry text and images across the open web. You deprioritize adtech. If you are ambitious you might try to build new AI applications that blend self expression, gaming and media. BuzzFeed took a swing at its first next gen AI experience. It's an auspicious start but not very entertaining. You definitely leverage AI to explore content creation efficiencies.
So pull way back to see things for what they are. The web starts to look different, half chat box, half vertical video.
The chat part shouldn't be much of a stretch. Truth be told, much of how we use the internet today is already a chat box. Google was the OG chat. We learned to ask it questions, it just returned structured answers in the form of web pages.
But AI is going to quickly evolve the structure of the well trod search-to-webpage interaction. Search finds. Chat extracts. Things change slowly then quickly and all of that. Open.AI’s GPT4 shows us that AI's progress is rapid, non-linear and now, alarmingly nuanced. The area underneath a search query is about to become much more personalized and far less structured. There will be endless pressure to accommodate content creators inside a new model, but the shape it takes and the rewards for media is unclear.
The arc of the internet is long and unpredictable but bends toward user empowerment and ever increasing fidelity. An endless stream of algorithmically sorted vertical video is the current endpoint. Robots that do much of the work to make sense of things for you are coming faster than you can say “human augmentation.”
But the bigger problem, one reflected in the BuzzFeed results and across the market is much of the revenue that underpins publishing businesses is going away.
The ad squeeze
There are essentially three broad types of digital advertising. Video advertising is simple and enduring. In short, it amounts to buying dedicated bandwidth inside your attention processor. Formats may get shorter or longer, it may be highly contextual or completely disconnected from the main act. These are questions that relate to strategy and efficiency and supply, the underlying advertising mechanic is the same and will always be so. Video advertising rents attention in exchange for the satisfying attention snacks you really seek to consume. It prices these by impression (CPM) which is essentially just a way to value attention. It is always in high demand.
For lack of a better term, let's call the next type "distribution advertising." Some people call it performance advertising. I call it "distribution" because that's essentially what it is. It extends the shelf space for a product of service to places where people are looking for the thing you are selling. As such it is just distribution masquerading as marketing. Search advertising is a good example. You search for thing. Advertisers buy space against this search surface area so you can more easily pick their thing. The internet made this kind of advertising possible because the internet connects intention to transaction. In so doing it has fundamentally changed the nature of marketing spend to more "cost of good sold" than a speculative marketing "investment."
Retail media is a subset of the above because it places your product in preferential places where people are shopping for stuff. It's amazing that it took so long to figure out that these positions (and associated data) are oceanfront real estate, but it's ascendancy was largely dependent on the maturation of e-commerce. Anyway, it seems to work very well, advertisers are hungry for it, it doesn’t require costly content... but its growth comes at the expense of our third category which we will get to in a second. To put retail media importance in context, Amazon, a leading purveyor of it, does more revenue in retail media than the entirety of the publishing industry.
Affiliate advertising is another subset of distribution. It places shopping links inside of the content where you are reading about things you want to buy, like a link to buy a TV at Best Buy inside of a review of “The best 65” TVs under $1000.” There are other, weaker forms of this like "retargeting" or “hyper-targeted” advertising but they basically just chase you around non-contextual environments because they know something useful about you and your intentions. These types of ads are almost always priced on a performance metric like cost per click, per lead, per sale, etc. Affiliate advertising is one of the few growth areas for digital publishers. We see this in BuzzFeeds results. The risk, of course, the affiliate incentive becomes too enticing, turning every media brand into Consumer Reports.
The third type is basically everything else you see online, let's just call it "opportunistic advertising". Its banners, content marketing and native ad units and such, the stuff that sustained many publishers like Buzzfeed. This category attempts to opportunistically bend your attention from one place and direct it to another. At worst it is noise, a flashing banner next to a news story. At best, contextual alignment or deep targeting pushes to approximate affiliate advertising above. The people who sell this want to price it like video, a function of attention. The people who buy it want to price it on performance. They later usually win.
Despite two decades of innovation in how best to present this kind of advertising in and around content, it remains the toughest to sell. Media businesses that rely on it have slowly come to realize that it's a weakened revenue stream best served by an efficient programmatic buying marketplace at very low yield. Only those with low-cost audience traps, scaled data, sophisticated ad matching algorithms and direct buying interfaces (Meta) are able to convert this rock to gold.
Video aligns interests through forced viewing. Distribution aligns advertising around user intention. The Opportunistic category lacks alignment and tries to manufacture it. Despite best efforts, this has not worked very well because it doesn’t align with either attention or intention. In hard economic times, it's the first to come under pressure.
The truth is, there was always a ton of fat and friction in digital ad sales. Selling the stuff led to swollen sales and marketing teams shuttling a lot of fancy slide decks with complex diagrams to apathetic young buyers at media buying agencies. In media, revenue on the margin flows to the bottom line. Everyone had the incentive to invest here, now it’s where you look to cut.
The Buzzfeed people are right to invest energy in creators and AI. This is where all energy is flowing. But navigating the new math will take dexterity: 1) Creators have much of the leverage; 2) Platforms that power them become even more important in a video world; 3) AI will do a lot of the work that media used to do; 4) Much of advertising increasingly looks like distribution and that’s hard for most to trap. For a few, AI may augment media teams in a way that yields new opportunities but how this creates lasting value is hard to discern from here.
Have a great weekend.../ Troy
ON THE PODCAST
Bank failure meets culture war
Brian shows his populist DNA. Alex defends the tech class. I mostly take it all in.
A pet-friendly pod. Best in media. Listen here.
But what should we call it?
Rozi Plain made a very nice song.
Rozi Plain was born Rosalind Leyden in Winchester, England, in 1986. In 2006 she moved to Bristol to study art, and she participated in the Cleaner Records group there. She recorded two records published by King Creosote’s Fence Records, Inside Over Here (2008) and Joined Sometimes Unjoined (2012). In 2012 she moved to London.
In 2015 Plain released a new album, Friend, followed in 2016 by a companion album of remixes, unreleased tracks and radio sessions, Friend Of A Friend. In 2019 she released What a Boost and in 2020 What a Remix.
She is a member of This Is the Kit and regularly contributes to the music of her friends, including Rachael Dadd, François & the Atlas Mountains, and Bamboo, with whom she has performed live on a number of occasions in 2016.