The Numbers Behind the Paywall

Posted on January 31, 2012


Finally! The New York Times is coming out with its paid-for content strategy. A quick summary of the Gray Lady’s paywall plan: a monthly allotment of stories to be read  for free and, above that, a flat fee for full access. Subscribers to the print version (including those who only get the Sunday paper) will have free access. According to the official press release, the new system will be launched in January 2011. For now, that’s all we know.

Why such weird timing? For the biggest online newspaper in the United States, announcing such a move just a week before the likely roll-out of the Apple Tablet is bizarre. OK, we get it, the New York Times won’t join other publishers aboard the Apple bandwagon. As I’m writing this, there are persistent rumors that big players such as Condé Nast, Harper Collins, McGraw Hill, Hachette could sign up — but not Time Inc., according to All Things Digital. Then, either The NYTimes is showing its fierce independence or it is hedging its bets by preparing its own offer, competing with a putative Apple publishing hub. According to New York Magazine, the Times is not joining Journalism Online (see previous Monday Note How to make readers pay for news), nor is it teaming up with the Wall Street Journal in its effort to pressure Google for a better deal. Another question: Why wait so long to deploy the NYT’s paywall? A year to build a digital subscription system sounds quite a long time.

Let’s throw some numbers at the “metered model” – as it is now referred to.

Which part of its audience does the Times actually target? Last November, the Times got 16,5m unique visitors and 2.98 sessions per month, according to Nielsen. As for the number of pages viewed by each visitor, we must rely on a more global figure, again from Nielsen: for the top US newspapers, an average of 43 page views per month (1).

The problem with the Times is we can’t go further down in the analysis because its quarterly financial statements don’t breakdown the economics for the company’s 50 online properties. Then, let’s turn to Washington Post. This newspaper releases numbers for its own web site, numbers we can use to try and model how a metered paywall.

For the full 2008 year, the made $122.7m in advertising revenue. Applied to a monthly audience of 11m unique visitors (we’ll assume this stable audience is a yearly average), this translates into $11 per visitor per year. That’s the ARPU (Average Revenue per User) for the Post, it comes only from advertising; and it’s 20 times less than the ARPU for print readers. You see the idea, the goal: having people pay for content is a way to close this 20X gap between ARPUs.

Now, let’s look at the problem from a different angle: who would be willing to pay for the Post’s or the Times’ prestigious content? Not that many. First of all, a large percentage of the audience lands on these sites though search engines and therefore are not likely to consume a lot of pages. Such visitors, about a third of the total audience, must be removed from the pool of readers likely to pay for content.

How much would people pay? According to a Boston Consulting Group survey, readers would agree to pay $3.00 per month on average. Interestingly enough, the BCG found the upper limit to be $6.00 for the “heavy print consumers” category. (The intentional amount is higher in Europe than in the US). As we explained in last week’s Monday Note (See The Death of Joe Average), the notion of mean value doesn’t make much sense here.

Coming back to the Washington Post, using the remaining 66% of total users likely to pay for content (7.34 mUV/month), the expected revenue impact is modeled below (in blue, the revenue boost when compared to the 2008 advertising only numbers):

Some will object: the site’s advertising revenue will be affected by the paywall. True and not true. Yes, the site’s overall audience (as considered in page views) will decrease slightly. But if the paywall only targets the viewership’s top 10%, the loss will be minimal. Second, experience shows ads placed behind a paywall yield much more money than those in the free part. The Wall Street Journal is said to charge 30% more for ads displayed in its pay zone. Advertisers set a higher valuation for this loyal audience.

In conclusion: a carefully set up paywall significantly increases revenue as long as:

  • it doesn’t block access for the general audience (minimum traffic loss)
  • it doesn’t discourage linking from other sites (preservation of the page rank)
  • it targets only the heaviest users, those willing to pay $6.00 rather than $2.00 per month, and those ready to be charged for ad-free content on mobile. Or on a Tablet.

( 1 ) Strangely enough, Nielsen/Mediametrie in France is showing only 11 pages per French reader of the online NY Times – OK, the language problem might lower the page count. But, for October 2009, (November data are not available down here, Gallic computers are slow, you know), Nielsen grants only 19 pages per person and per month for Le Figaro and a mere 15 pages for Le Monde. These two big sites are comparable by their scope to the NYTimes or to the Washington Post. How come French readers view half as many pages as their American counterparts do – for the same type of sites (if I may ask)? — An update: the cause of the delay to get the French Nielsen data is said to be bug in the US. As for the discrepancy in the page count between US and French, I’m on it.


Posted in: News-Trend