How the NYT paywall could turn out to be a success

By Felix Salmon
March 23, 2011

There’s one aspect of the NYT paywall which hasn’t got as much attention as it deserves — and that’s the idea that the NYT will be able to charge higher CPMs for ads behind the paywall than it currently gets for ads on a free site. Gordon Crovitz, the former publisher of the WSJ who now has a paywall product of his own, points out that if people pay for a subscription, that’s an excellent indicator of the “engagement” that advertisers and agencies are looking for. As a result, he says, publishers can charge a CPM premium of about 30% for behind-the-paywall pageviews.

Of course, it’s hard to know whether the NYT can get that kind of premium. At the NYT, after all, most behind-the-paywall pageviews will come from people who aren’t paying for a digital subscription: either they’re print subscribers who get access to the website thrown in for free, or else they’re online-only readers who have taken advantage of the Lincoln offer. Do people who get free access behind the paywall count as “engaged” from an advertiser’s perspective? I’m sure the NYT sales team will do its best to persuade media buyers that they are.

What’s more, if those people are told clearly that they’re getting something worth hundreds of dollars for free, then they might end up using it more. So even if CPMs behind the paywall don’t go up a lot, it’s reasonable to assume that readers with a get-past-the-paywall pass will end up reading more NYT stories than they have done up till now.

Against that, of course, is a certain drop-off in traffic from readers who don’t want to pay for access, who haven’t taken advantage of the Lincoln offer for whatever reason, and who have no desire to engage in clever tricks any time they want to read a NYT story.

It’s all a lot of swings and roundabouts, but think about it this way: let’s say that 2 million web denizens have access behind the paywall, by the time that the NYT is done with its introductory offers, multiple accounts for home-delivery family members, and the like. I’m running about 150 articles a month, but I’m a heavy user; let’s assume that on average those 2 million people read 75 articles per month. That’s 150 million articles per month, and with some articles containing multiple pages, let’s call it 200 million pageviews per month, or 2.5 billion pageviews per year.

How much extra money could the NYT get from those 2.5 billion behind-the-paywall pageviews? If it manages to raise its revenue per thousand pages (RPM) by $10 for pages behind the paywall, then we’re talking about an extra $25 million in ad revenue — which compares to total digital ad revenues running north of $300 million per year.

It’s a big if, of course — the digital account for the family 9-year-old is a nice perk to keep the print subscription coming, but it’s not going to attract a lot of advertisers. But if the NYT can make the sale, then that $25 million could go a long way to helping to make up for any reduction in total pageviews which comes from introducing the paywall. At that point, the subscription revenues don’t need to offset a decline in ad revenues: they just need to pay for the cost of building the paywall in the first place, help to diversify the NYT’s income streams, and encourage readers to buy the NYT in its highly-profitable print form.

The point here is that the paywall is responsible for multiple revenue streams, not just its own narrow subscription revenues. It will drive marginal readers to the print product, especially at weekends — and thereby shore up or even increase print-ad revenues from the weekend paper. It will dissuade current print subscribers from dropping the paper on the grounds that they can get all the same content for free online. It will allow the digital ad-sales team to charge a premium for readers who have access behind the paywall. And, as we’ve already seen, it will allow deals where advertisers like Lincoln pay good money to sponsor that access.

Add them all up, and I’m beginning to come around to the idea that the paywall can make good financial sense — if everything goes according to plan. This is a complex system, and therefore prone to misunderstandings and mistakes: there have been predictable glitches in Canada already, and I was talking to one NYT staffer on Monday who was convinced that all of the blogs were free, rather than just the “blog fronts“. But over time I suspect that the NYT will manage to simplify both the system and the messaging, which Ken Doctor points out has proven rather tough and unfriendly at launch.

The NYT will surely proclaim the paywall to be a success no matter what happens. But if total pageviews don’t fall and digital advertising revenues increase, then it’s going to be pretty hard to make the case that the paywall was a bad idea.

If the paywall does succeed, I’ll be very happy indeed to have been proven wrong — the NYT is a great newspaper, and it deserves some financial good fortune. It’s a real possibility; let’s hope it becomes a reality.


We welcome comments that advance the story through relevant opinion, anecdotes, links and data. If you see a comment that you believe is irrelevant or inappropriate, you can flag it to our editors by using the report abuse links. Views expressed in the comments do not represent those of Reuters. For more information on our comment policy, see

Couple things. If I paid for access, I will find ads offensive, what the heck am I paying for? I’m not just paying for access, I’m paying to not be bothered by ADS!

Second thing, there is no standardized accountability when it comes to internet advertising. Even the most ethical standards put in place will suffer from those who stack their stats.

here are some examples of stat stacking that I think I have stumbled upon as a browser. I’m following 60 second game updates on any of a number of well known sports platform. Everytime the game update refreshes, a new ad appears, and the website can count that as a new ad hit! Well, my question is, just when does a viewer reach saturation on a site?

If I am watching for game updates, and in the course of an hour I allegedly see 60 different ads, did I really notice more than a handful of them?

Some websites with forums and comments sections force their readers to relog in every few hours. Sometimes the loyal reader/viewer has to go back to a main page to re-sign in, and then navigate through a few links to get back to where they were. Are these all considered separate page hits?

Another annoying tactic involves loading multiple ads at the same time that use too much memory and clog up the actual loading of the page.

Other annoying ad tactics are promising some type of information or item for “free”. but then not allowing access to said free item or info until the reader has given up their own email address and personal mailing address as well.

The sad truth about the internet is that most people like to flit from site to site, sometimes even establishing a routine, and they just have no desire to buy anything. The result is incredibly low ad response rates.

Posted by SWARMtheBANKS | Report as abusive

One pays for the print edition, and gets ads. I don’t mind ads, as long as they are not the intrusive “skip this ad and go on to the article” preview type. Who knows, the ad might actually be something I’m interested in.

I don’t mind paying for the Times, (although I did accept the Lincoln offer). Newspapers aren’t selling paper, they are selling news. They have to pay investigative reporter, and to do that they need an income.

Posted by Loebner | Report as abusive

Sorry, I thought you knew about the difference between paid subscribers and free copies. This is a key of the newspaper model. Newspaper subscriptions were carefully audited, separating paid subscriptions from daily sales from freebies.
And I think you are describing part of the Times strategy, but I think their motivation lies elsewhere, if they’re smart (which is hardly a given when you are talking about newspaper companies.)
Paid subscribers back in the day were worth more because those people were motivated enough to fork over dough to get the paper regularly. The assumption is, as you say, they are looking more closely at content than the free readers.
But that metric only makes sense in the old media, when it was possible to give away, oh, 100,000 copies of a paper but never really know how many people took the time to open it. Then advertisers had to gauge the likelihood of seeing their ad, given that a paper had found its way into the hands of a reader. They had to measure somehow that the guy with the paper actually read it, instead of just throwing it in the trash.
But paid circulation is not a precise metric. Lots of people used to subscribe to a paper and let unwrapped papers pile up just inside their doorstep. They’d only unwrap the papers to prepare them for recycling. However, it was the best measurement device available.
With the internet, paid subscription remains a weak gauge of whether subscribers were exposed to an ad. You don’t need an indicator at all. You can find out how many times your ad was placed on a page in front of people. There are already tons of more sophisticated metrics.
Example: Reuters doesn’t need a paid subscription base to tell advertisers how much people like Felix Salmon’s blog. (Right now ING has the banner on my screen.) They can report that X people looked at Felix’s blog on Wednesday, March 23, and Y of those people were exposed to the ING ad (I’m assuming the Felix Salmon banner is sold to more than one advertiser.)
In fact, paying a premium for paid subscribers doesn’t make a whole lot of sense, since along with knowing who clicked, you can describe a whole lot of demographic information about the clickers.
There are two possibilities.
First, advertisers, like the Times, haven’t really thought through the stuff I just wrote about. You can’t discount that possibility, because the idea of CPM has a century of momentum behind it, and the people who run businesses can be victims of that kind of statistical tyranny. (Separating daily and Sunday circulation had the same effect on the newspaper, but that’s a story for another day.)
Or the Times may be cleansing its demographic. It can present the average paid subscriber as wealthier than the average Lincoln-fed subscriber and charge a premium to appear on paid subscribers’ pages. This makes sense if you figure the average Times paid subscriber is a wealthy person to whom $15 a month is meaningless.
So the paid pages would have lots of ads from Lexus while the Lincoln-fed ads would have lots of ads from, say, Wal-Mart.
My example, though, kind of shows how shaky this thinking is. Why would Lincoln want to sponsor the Wal-Mart pages?
I’d love to see the demographics of the people Lincoln is offering subscriptions to. I suspect it has a lot more to do with median income than with account activity.

Posted by RZ0 | Report as abusive

@Swarm: We never had cable growing up, and at some point in the mid nineties I remember being shocked to discover that cable had ads. People *pay* for cable; it shouldn’t *need* ads. Why do people pay for it if it still has advertising?

Felix: suppose advertisers are rational. (Just for the sake of argument.) It seems likely to me that whatever benefit they’re drawing from selling the advertising is well approximated by a sum of benefits associated with ads served to each viewer — 42 cents from dWj seeing our ads this month, $3.24 for Felix (since he’s a big spender), etc. — with higher numbers per pageview basically represent higher amounts of “engagement”. If we remove the lower-valued pageviews, the average will go up — say by 30% — but the total will go down. I would think advertising revenue would ultimately be reflective of this; the only way you end up with more advertising revenue is if you can draw in customers (or engagement) that wasn’t there before.

Perhaps, as you suggest, even if the advertisers are rational, the readers aren’t — they’ll take their free pass from Lincoln (or a paid subscription, for that matter), and tell themselves how much more they value NYT than they would any free website, and they will provide the extra aggregate engagement there. General equilibrium is complicated, and perhaps something else would turn up — maybe the comments on articles improve, and draw readers who otherwise wouldn’t have been there, or the selection of articles starts to change. Most of the stories I tell myself that seem likely to lead to an increase in ad revenue, though, require new readers. I think the first-order effect is a decline, avereage engagement level notwithstanding.

Posted by dWj | Report as abusive

Forgive me if I am missing a point here but printed newspapers also have adds in them. Even if a person does not subscribe they still have to pay for it on an individual basis. So what’s the difference if the website also has ads even if one subscribes to it? I mean SWARMtheBANKS finds it offensive to have ads on a web site paid for to access. Is SWARMtheBANKS also offended by ads in printed newspapers of which are also paid for?

Posted by iflydaplanes | Report as abusive

For those making comparisons between print ads, consider the differences as well.
* I’ve never had a print ad crash my newspaper and force a browser restart.

* I’ve never had a print ad throw itself on top of the article I’m trying to read and insist that I find the “X” if I want to go on reading.

Perhaps the NYT has a higher class of ads than their subsidiary, but the Boston Globe is increasingly annoying to visit. The result, naturally, is that I’ve mostly stopped reading it. When I do, I navigate directly to the blogs or features that I want and avoid the aggressive main page.

I would pay for the Boston Globe if it eliminated the aggressive ads. But (as my viewing habits have shown) I will leave an ad-ridden website for free.

Posted by TFF | Report as abusive