Running the numbers on the NYT paywall
Apologies for being something of a one-trick pony today, but reading John Gapper’s column on the NYT paywall makes me realize that a lot of people fail to appreciate exactly what’s at stake here. Gapper seems to think that online subscription revenues can make newspapers profitable again; they can’t. In fact, insofar as the paywall makes any sense at all, it does so only as a tool to boost print subscriptions.
Gapper notes that the Guardian’s parent company lost $93 million in the last fiscal year, despite having a website attracting 35 million unique visitors globally, and 13 million domestically. He reckons that charging some of those subscribers could make the Guardian’s problems disappear:
Outsell, a research group, reported this week that only 6 per cent of US online readers say they would pay online news sites if they charged.
If we are to take the figure at face value (which I don’t think we should), then The Guardian could get 2.1m people to subscribe to it online, making it highly profitable at a stroke.
The logical flaws here are huge. Let’s just count a few:
- Gapper is assuming that if 6% of online readers would end up subscribing to some site or other, that’s the same as saying that any given site can persuade 6% of its readers to subscribe. But that’s not the same thing at all.
- Gapper is taking a poll of US internet users, who are more likely to pay for things online than other nationalities, and extrapolating the numbers globally.
- Gapper is assuming that foreign readers (say, Guardian readers in India) are just as likely to subscribe as domestic readers (Guardian readers in the UK).
- Gapper is assuming no overlap at all between the 6% of readers who would pay online, and the percentage of readers who already pay for a newspaper subscription in print form. The NYT has already said that its print subscribers will get an online subscription for free, and I’m sure the Guardian would do likewise.
- Gapper is ignoring that putting up a paywall will always reduce advertising revenue, which means that in order for the new subscription revenue to make the newspaper “highly profitable at a stroke”, it would have to not only make up existing losses but also cover the drop in online advertising revenue.
Once you understand what Gapper can do to statistics like these, then, it becomes a bit easier to make sense of something like this:
Rates for online display ads have been falling steadily as competition has proliferated, with most sites now finding it hard to get more than $4 per 1,000 impressions on their pages (or $14m for the 3.5bn hits on all US newspaper sites monthly).
I have no idea where Gapper’s getting his $4 CPM figure from, but it’s clearly much closer to being a minimum than an average. Papers might find it hard to get more than that, but you can be sure that the NYT, for one, is succeeding all the same. In fact, the WSJ reports today that nytimes.com is pulling in $100 million in revenues annually — more than you might think the US newspaper industry as a whole was making, if you read Gapper too literally. Indeed, in the third quarter of 2009, the New York Times Company made $79 million from its internet businesses, of which $68 million came in advertising revenues. Newspaper advertising was $39 million — and that was in a very weak quarter. On an annual basis, I’d be surprised if nytimes.com didn’t make significantly more than the $100 million that the WSJ is talking about.
It’s also worth noting that Gapper has managed to confuse CPMs — the amount of money that an advertiser pays per 1,000 pageviews — with RPMs, or the amount of revenue that a publisher receives per 1,000 pageviews. There’s nearly always more than one ad unit per page, which means that RPMs are some multiple of CPMs, depending on how much of a newspaper’s inventory is sold.
Gapper, then, is systematically overestimating the upside of subscription revenues, while underestimating the magnitude of advertising revenues. Erick Schonfeld, by contrast, is much more realistic, concluding that total subscription revenues from nytimes.com would optimistically reach only $9 million per quarter, or $36 million per year. With the New York Times Company making the best part of $300 million a year from online advertising, it’s hard to see that the extra revenue boost would really be worth it.
The point here is that with the powerhouse nytimes.com site front and center, the New York Times Company as a whole is a major online media player, serving up billions of high-prestige pageviews and building strong relationships with every major online advertiser and media buyer in the country. Even under the most optimistic scenario, a majority of the NYT’s loyal readers will desert it when it moves to a paywall. And with those readers gone, media buyers are by no means guaranteed to stick around.
Gapper makes great play of the fact that websites can target ads more accurately when readers are registered, but you can’t target ads at readers who no longer exist. And the NYT is a mass-market general news publication: it’s not the kind of place where high-end business-to-business advertisers will pay $90 CPMs to reach C-suite executives. Or if it is, the numbers involved would be so small that they wouldn’t make a visible dent in its overall online advertising revenues.
What’s a realistic number for how many people will pay to subscribe to nytimes.com? David Carr says that the NYT wants to target “10 percent or so” of the 17 million current readers of nytimes.com. That’s 1.7 million people. Subtract the print subscribers who will get nytimes.com for free, and you’re left with 1 million, more or less. How many of those could you dare hope to persuade to subscribe? One third? Once again, just as Schonfeld did, we get to somewhere in the region of $35 million a year, assuming a subscription price of $100 each per year. For a company with annual revenues in the billions, the hit to the value of the brand alone has to be bigger than that.
There is one other dynamic at work, here, however, and that’s the price of the print subscription, which has proved to be surprisingly inelastic: David Carr in fact lauds the way in which “The Times has shown a great ability to leverage prices once they have custody of a consumer”. But pushing existing consumers to the limit of what they can pay only makes it that much harder to attract new ones.
The NYT aspires to be a national paper, which makes sense, since it either already has or is never going to get most New Yorkers as print subscribers. But the annual subscription rate, if you live at say 1600 Pennsylvania Avenue, is $769.60, and it’s really hard to get people to pay that kind of money, in the middle of a recession, for a newspaper they’ve lived happily without for all their lives, and which they can get online for free.
So I see the decision to implement a paywall not as an attempt to build a significant revenue stream for the website alongside ad revenues, but rather as an attempt to shore up — and maybe even increase — the print subscriber base. Even a Monday-Friday subscription, at $384.80 a year, brings in much more money than any online subscription ever will. Yes, it probably costs more than that to print and deliver the paper. But for the time being at least, print advertisers are still willing to pay top dollar for a full page in the physical New York Times. And so long as that’s the case, the NYT will do anything to keep its physical circulation numbers as high as it can — even, it seems, if that means dealing a serious blow to nytimes.com.
Update: Gapper responds with a 950-word blog comment. Good for him!