Content economics, part 3: costs

When I wrote last week about Jeff Bezos and his journalists, I said that “the Boston Globe was sold for essentially a negative sum, once pension obligations are taken into account, while the Washington Post was sold for the price of a nice Cézanne.” It turns out that I was comparing apples with oranges: the Washington Post, just as much as the Boston Globe, was sold for less than the value of its pension obligations. Bezos might have paid $250 million for the paper, but he was also given $333 million to help him meet its pension obligations.

Make no mistake: this does mean that the Washington Post is worth substantially less than zero; you can’t just separate out the pension obligations and declare that somehow they don’t matter. What Bezos paid for the Post was not $250 million, it was negative $88 million.

To be sure, Bezos is $250 million out of pocket: the $333 million in the pension fund is ring-fenced for the Post‘s pensioners. But think about it this way: what Bezos bought was the institution of the Washington Post, the brand which the Graham family spent decades investing in and building up. Bezos bought an established property, because it was an established property. If he didn’t want an established property, he could have invested $250 million to create a brand-new journalistic outlet. To put that sum in context, the total amount of money raised by Business Insider, since inception, is $21.6 million; Vox Media has raised $23.5 million; BuzzFeed is on $46.3 million; and Huffington Post raised $37 million before it was acquired by AOL. Throw in Gawker Media, Mashable, Politico, Pando Daily, Breaking Media, Weblogs Inc, and just about any other new journalism company you can think of, up to and including Bustle; you’re still nowhere near $250 million.

Why does it makes sense to buy the Washington Post for $250 million, rather than spending $250 million creating an awesome journalistic product from scratch? Because the Washington Post is already established — and while it’s easy to buy journalism, and even traffic, for $250 million, it’s harder to guarantee yourself the kind of name recognition and national reputation enjoyed by the Washington Post. That reputation is built on hard-won promises: promises to readers that they can trust what they’re reading, and promises to writers too. The deal that the Washington Post made with its writers — the deal which helped to cement its reputation — involved not only paying them a present salary, but also promising them a decent pension. That promise helped to build the brand, and it was attached to a fully-funded and very well managed pension fund. The pension fund is, in all senses of the word, an important part of the value of the Washington Post. And Bezos just managed to acquire a $333 million pension fund, which only has about $283 million of liabilities, for $250 million. Which says to me that the value of the newspaper itself is clearly negative.

When any institution has a negative value, that’s because it’s likely to lose money over the long term. And there’s a good reason why the Washington Post is likely to lose money over the long term: it’s a large journalistic organization. And the internet has turned the economics of journalism on its head: once upon a time, profits and size went hand in hand, with the biggest outfits making the most money. Today, it’s the other way around: small, lean companies make modest profits, while bigger outfits generally see ever-increasing losses.

Part of the reason is related to revenues. I argued in part 2 of this series that if you want your readers to pay you, it often helps to be small and personal rather than big and corporate. And I argued in part 1 that the economics of advertising were not favorable to news sites — while advertisers, once upon a time, might have felt as though they had no choice but to appear in a big publication like Time or the Washington Post, nowadays the sites with the biggest traffic aren’t journalistic at all.

But there’s a key part of the economics of journalism which I haven’t covered yet — and that’s the question of costs, specifically the costs of producing the stuff. On its face, journalism is a classic scalable commodity: once I’ve written this post, for instance, it can be read by a million people as easily as it can be read by ten. And so if it’s published by a big journalistic organization with wide reach, ideally one where a substantial and talented business team can help build lots of pageviews which they then sell to deep-pocketed advertisers, then a fixed cost (me) should be able to support lots more revenue than I could at a smaller site.

Certainly this is true narrowly, at the margin: the economics of online publishing involve fixed costs and variable revenues, with the result that all things being equal, any given site will make more money as it gets more pageviews. But all things are not equal. And if you start looking up the spectrum from smaller sites to larger, more established organizations, instead of simply looking at what happens to any given site over a relatively short time frame, you find a very different pattern.

If you look at the media world as a snapshot, instead of a single site over time, it turns out that editorial costs rise steadily as sites get bigger and more professional. They rise in absolute terms, of course. But they also rise in per-editorial-employee terms, and they tend to rise even in terms of editorial costs per pageview. The lowest costs come where everything is written and published by individual bloggers — think any number of unpaid blogs, or, perhaps, think of the blog network that is Glam Media. The highest costs, meanwhile, come at the large outfits which employ hundreds or even, in some cases, more than a thousand journalists.

Back in Henry Luce’s glory days, big publishers could easily absorb spiraling editorial costs because there was so much money at the end of the rainbow: once you achieved a certain level of circulation, you basically became a license to print money. The online world, by contrast, has no magical rainbows: no matter how big a news site becomes, it will never be so big that advertisers will clamor to appear on it, whatever the cost.

The result is that the journalistic outlets seeing the biggest profits are the ones where costs are kept incredibly low. At one extreme lies the Bleacher Report, which was sold for a reported $180 million; it consists primarily of stories written by unpaid contributors, expertly optimized to maximize pageviews at the expense of accuracy or quality. Or look at Summly, which Yahoo bought for $30 million: it produces news summaries entirely by algorithm, with no editorial costs at all.

And if you look at most new journalistic websites, you’ll see they have a few things in common, many of them very admirable. They move fast, they publish in very high quantity, their staff journalists are extremely prolific, there’s relatively little editing, and they’re happy to rely in large part on reporting done elsewhere. That’s the kind of thing the web does very well — and it’s cheap to produce. Most online journalistic organizations were founded after blogging tools became available online for free around 2002; the few which predate that, like Salon, Slate, and The Street, have tended to struggle with their costs and profitability.

This is not to say that the web doesn’t have wonderful expensive content. Investigative work, longform narrative journalism, beautiful immersive multimedia experiences — we’re seeing it all. But that kind of stuff costs real money: ProPublica, for instance, spends about $10 million per year to support 19 reporters, plus roughly the same number of editors, executives, developers, and support staff. That’s over half a million dollars per working reporter per year, in a place where stars like Jesse Eisinger make more than $200,000 just in salary. That’s not the kind of thing you’ll ever see at, say, Gothamist, where the editor of DCist, who was recently fired for writing a freelance piece for BuzzFeed, was being paid $40,000 per year.

There’s a good reason why Eisinger is being paid so much money at ProPublica: it’s easily what he’s worth, on the open market. His skills and expertise are in high demand among ambitious, high-quality news organizations — and not because he drives millions of pageviews. At the same time, those skills also require diligent editors with many years of experience: investigative reports are, always, an expensive, time-consuming team effort, and an effort which sometimes needs to be aborted after many months and enormous investment.

Here’s a statistic worth dwelling on: “A senior editor at The Washington Post recently told me that he killed an average of three advanced investigations a year, usually over the protests of the reporters, who couldn’t see that they didn’t have the goods.” Outside ProPublica — and even inside it — how many online-only organizations can say the same?

The stat comes from the enormous meta-investigation by NPR’s ombudsman into an investigative piece about Native American adoptions. The amount of time, effort and money that NPR invested both into the original report and then into the ombudsman’s report is a good indication of just how expensive journalism becomes, when it takes itself seriously and has the highest ambitions. When Jeff Bezos bought the Washington Post, he bought an organization which spikes three advanced investigations per year. That’s not efficient, or cost-effective, but, whether he knows it or not, it’s part of the reason why he liked the Post enough to buy it.

When I say that “greatness emerges mysteriously from the slack in the system”, this part of what I’m talking about — a system where expensive projects regularly get the green light, even if they’re not shoot-the-moon ambitious. The lower you set the bar, the more failures you get — but also, the more serendipitous successes you find.

Last month, for instance, the New York Times published Kate Taylor’s story about sex on campus — an article which was met with much derision in the blogosphere, and which seemed to add little to what the NYT had already published on the subject. This article, like its predecessor, appeared in the Styles section of the newspaper, and the “Fashion & Style” section of the website. These are sections which exist only because of advertiser demand, and are not taken particularly seriously within the newsroom. And yet, it turns out, Taylor worked on the piece exclusively for five full months from September through January — and then “on and off” between February and July. At the NYT, even the Styles section is willing to invest what must have been close to an entire person-year, once you include the time of the other people who worked on the piece, into a single 4,800-word article. Whatever you think about Taylor’s piece, that speaks volumes about the internal culture of the NYT, and the costs involved in producing it.

In her open letter to Jeff Bezos, Kara Swisher applauds the fact that the Post‘s new owner has “the gazillions of dollars needed to pull the Post through to the other side and allow it to maintain its top-notch standards in a world still unwilling to pay for some of those standards and the fine journalism that results.” The alternative can be seen at Forbes, where Lewis DVorkin, who revels in the title of “Chief Product Officer, Forbes Media”, recently announced that he had reached a big milestone — 50 million unique visitors in one month. This was reached by using a small number of editors to oversee a network of 1,100 contributors, only 25 of whom (2.3%) made more than $35,000, and only two of whom (0.2%) made more than $100,000. That model, he said, needs to “evolve”:

The digital space continues to move fast. Programmatic buying of ads continues to put downward pressure on ad rates. Soaring mobile usage creates its own challenges. Nearly 35% of our audience accessed through smartphones and tablets last month, up from 25% a year earlier and 10% a year before that. Mobile screens don’t produce the same revenue as desktop screens, nor does international traffic. We’ll continue to work with contributors as our partners to adjust the payment model to meet the economics of the marketplace.

Translation: We were paying very little before, and we’ll be paying even less going forwards.

The Forbes model is what happens when you take a legacy media brand and try to make it work in the context of the pitiless economics of the web. That’s clearly not what Bezos is going to want for the Washington Post — if it was, I doubt very much that the Graham family would ever have sold him the newspaper at all. But the alternative, given the size and journalistic importance of the franchise he just bought, is going to be much, much more expensive.