So I was very interested to see what seemed yesterday to be a promised vivisection of, well, me, after my post last week about the collapse of programmatic advertising and how it’s affected the news business. That is the post that included the chart of rapidly declining programmatic ad revenues. Gawker alum Foster Kamer linked to a promised takedown from Ben Smith in which Kamer is quoted and calls on everyone to stop “losing their minds at the death of media” because it’s a “very dumb chart.” (That’s yours truly’s chart: cue to Josh rending his garments.) But when I looked at the Ben Smith column it basically just repeated the points I made in the original post. So having been promised that I would be flayed alive, it turned out that wasn’t the case at all. Total low-energy move basically. But the comments did raise a few other issues that I thought worth mentioning.
I’m reposting the chart here just for the purposes of illustrating what I’m discussing.
![](https://talkingpointsmemo.com/wp-content/uploads/2024/03/Screenshot-2024-03-06-at-9.35.37 AM.png?w=804)
As I made clear in the post, the chart tells part of the story of one company’s successful efforts to grapple with the vast changes in the programmatic and larger advertising landscape over the last decade. On an apples-to-apples basis, revenue from programmatic advertising hasn’t gone down by 95%. This steep drop off is also due to changes we made in reaction to the cratering of the programmatic ad market. Our traffic is lower because of the changes in social media, our limited use of a paywall to sustain our membership business and the fact that we now have a membership tier that has no ads at all. Each of those developments is either tied to the evolution of the major platforms (Google, Meta, et al.) and their role in the ad business or the ways we have changed our own model in reaction to them.
But we didn’t just decide this was money we didn’t need anymore. The changes we made that played a direct role in the decline were entirely in reaction to reductions in potential revenue which we knew we couldn’t sustain. While we were making those changes we still fought for every dollar we could get out of the rapidly diminishing programmatic advertising pie. The results are what you see in that chart, which not surprisingly got a lot of people’s attention.
After I published the piece, I tried to think what would that number might be if we had done nothing but try to hold on to as much of that programmatic pie as possible. So no membership business, no paywall, no ad-free version, no shift in how we prioritized page views. The main answer is that we’d be out of business.
But still, it’s helpful to know as well as I can answer what an apples-to-apples comparison would be. I gave this some thought and talked it over with my colleague who I’ve worked with on this for the last decade. And I think the best answer is that it would be a struggle to keep that number at a third of what it was in 2016. Maybe a bit lower? Possibly a touch higher, but I doubt it. The truth is that it is very hard to arrive at anything concrete because we have changed every aspect of the organization in line with a very different business model. It’s not just a matter of fill rates and page rates and paywalls. These changes go deep into the editorial process as well, how we train up reporters, what we have them focus on. Needless to say, no company can withstand a 2/3rds drop in a primary revenue stream.
Kamer says this is all sky-is-falling hyperbole. In fact, programmatic advertising revenue has gone up, as various adtech influencers are happy to tell you. Kamer’s site Futurism is thriving on programmatic revenue, he says. It’s totally doable. In fact, he says I’m stifling innovation with all my nay-saying.
I have a few thoughts on this.
The first of which is, who are we trying to kid here? Does anyone think that advertising — direct or programmatic — still sustains digital news organizations, especially independent ones? Really? I think the almost weekly lists of bankrupt and shuttered news outlets tells the story pretty clearly.
But I’m not trying to get into a fight with Kamer. It’s awesome that his publication is thriving. Genuinely. I’m a friend to all cash-flow-positive media businesses. There’s a huge media ecosystem out there and there are certainly niches and whole areas where organizations can thrive. I applaud the ingenuity.
But there are a few points of context that are important to note here.
We should be clear what we mean by “news.” Tech news, fashion news, entertainment news are just as much news as the news we do. But they are very different in business terms. Those news spaces have endemic advertisers — gadgets, clothes, movies, etc. That’s a big difference. There are no industry sectors for cultural polarization and societal decay, where we operate. They also don’t face the negative premium that news publishers — in the sense of news about daily events and politics — face in a polarized age. Futurism focuses on tech and science. That’s a significant difference.
[Side note: There’s one part of the ad market where politics news is money: in the D.C. metro and what we might call para-D.C. That is essentially a subset of the corporate America’s lobbying budgets. That’s a separate issue I’ve addressed in earlier posts.]
Another interesting detail comes in this exchange spawned by Ben’s initial post and one of the guys he quoted. It’s about the start-up world for programmatic ad sales and ad tech that I mentioned in my initial post. A guy name Chris Tolles notes having been part of that gravy train I discussed in the first post and seeing it collapse. Ryan Brown, a Hearst Exec, said that what he wasn’t talking about didn’t count because it was “remnant” and not “programmatic.”
This gets very in the weeds. But it’s worth digging into it. Because there’s something interesting down there in the weeds.
Back a decade ago, and longer, most news publishers thought in terms of direct advertising and “remnant.” Basically, you sell as much as you can to high-dollar premium advertisers in the old-school way and then what’s left over you sell for pennies on the dollar as “remnant.” One of the ways I built TPM was realizing very early on that if you could come up with really creative ways of slicing up the ad inventory pie you could make decent money selling most or all of your inventory that way. This is how I grew the company in the critical years from 2005 to 2009 and then again along with a substantial direct sales business from around 2012.
But this is more a semantic point than a real one. What’s “programmatic” and what’s “remnant” and why should the distinction matter? I along with probably nine other people know the distinction that Brown is drawing. But again, what is the significance of that? “Programmatic” is a technical process. “Remnant” refers to what was both a strategy and a tier of the ad market. At the most basic level what programmatic means is the sale of ads through real-time digital marketplaces where advertisers “programmatically” bid for inventory. In the old days there was high dollar “direct” advertising and the lower-tier sales that were generally handled “programmatically” or often in more jury-rigged ways, splitting the inventory between different ad networks.
So high dollar direct sales and then a second tier in which “programmatic” and “remnant” were often used interchangeably as labels. The key in the latter category was that everything operated on Google’s infrastructure. And the secondary players mostly got absorbed by Google or run out of business by them. The start-up backed ad networks that we were slurping up money from as they fought to buy up market share withered away. There was an intermediate period in which those ad networks operated within Google’s programmatic bidding system. The key to understanding what was happening here — and why “programmatic” is such a wiggly descriptor — is that a lot of what used to be the high dollar direct advertising got routed into the programmatic Google-run infrastructure. So when you see ad industry people saying, “Wait! No! Programmatic spending is actually going up!” Well, sort of … maybe? Who knows, really? Not my monkey, not my circus anymore. But when they say that programmatic is “growing” they’re really just saying it’s gobbling up what remains of the ad market. It’s misleading at best, though I’m not sure how many people have long enough institutional memory to even know that.
Let me conclude on a more general point. We’re not complaining. For us as a company, this isn’t a sob story. We’ve built a new and very different business model that we’re happy with. We owe it all to you, the glorious TPM Reader-Member who will be honored in the halls of the gods as long as there is an internet. Our total revenue is at least in the ballpark of the highest its ever been and we hope to grow that and add more of the coverage we think you love. Also, to the extent we’ve reacted to dramatic downturns driven by the platforms, we benefited from earlier iterations of those platforms in the first decade of the century. Hard to say whether it’s one for one. My point is simply that as a business we are a boat on a vast ocean grabbing every opportunity we can and reacting where we must. But this perch does allow me and has allowed me for many years to have some insight into the broader news media business ecosystem. In some ways being so small provides unique insights because size and money can allow you to ignore many of the details. They give you lots of time to pretend they won’t end up applying to you. Until they do. Think of scale as a mix of morphine and trust fund. It let’s you ignore quite a bit.
My aim in these occasional interventions is to illustrate what other companies either can’t or won’t. Which is basically a look at why all these companies keep going under. Why did The Messenger fail? Well, these are the reasons. The funding base of the digital news space as it existed for the first 15 years or so of the century just doesn’t exist anymore, not at the levels that can sustain the labor intensive work of real political news journalism.
And as I said above, really, who are we trying to kid here?