A Newspaper Pay Wall Goes Up–And So Do Visitor Numbers

The New York Times is getting ready to roll out a “metered model” pay wall in January, and plenty of people fret that the paper will see its audience disappear when the (porous) gates go up. Here’s a counterargument: The Telegram & Gazette.

In August, the Worcester, Mass., paper put up a Times-style pay wall: Visitors can read 10 “local” articles a month for free, but after that they need to pay up. It’s not a coincidence that the Telegram is using the same idea that the Times will try in a few months–the paper is one of several local titles owned by the Times itself.

So. How’s that Telegram doing since the wall went up?

Just great, Times CEO Janet Robinson said during the paper’s earnings call today: The Telegram’s metrics are “on plan,” and traffic hasn’t suffered.

In fact, Robinson said, the Times was pleasantly surprised to see that the Telegram’s unique visitors number had increased since the wall went up.

What’s that? I asked the Times for numbers to flesh that one out, but it declined. ComScore, though, does back Robinson up: The Web traffic counter says 281,000 U.S. unique visitors came by the Telegram in August, and that number crept up to 294,000 in September.

That’s a tiny bump, though comScore often has a difficult time measuring smaller sites. For comparison’s sake, note that the Telegram tells advertisers it reaches 700,000 uniques a month.

Still, a bump is a bump. And it’s certainly not the plummet that many people would expect. When the London Times put up a pay wall this summer, for instance, it saw traffic drop a reported 90 percent. (News Corp. owns both the London Times and this site.)

So how do we explain the Telegram’s increase? In the absence of input from the Times or the Telegram (I’ve asked both for comment), we have to speculate. Feel free to add your own in, but I can start with a few theories:

Maybe the Telegram had some particularly blog-friendly, Facebook-friendly or Google-friendly stories in September. If that’s the case, the metered model would work well for the site, since it encourages casual visitors to show up via referral, without having to pay up. For a relatively modest site like the Telegram, you wouldn’t need many high-traffic stories to push up its base number.
Or maybe it’s just as simple as a seasonal spike: Traffic numbers often droop in the summer, when people have better things to do than sit in front of their browsers, and then spike back up in the fall.

In any case, this should give the Times a bit of confidence about its strategy for the flagship paper, which it promises to tell us more about soon.

UPDATE: Some readers are having a hard time accepting Robinson’s assertions and comScore’s numbers.

I have no reason to think that Robinson, the Times or the Telegram made the data up. If you’re a conspiracist who thinks otherwise, you should note that the NYT wasn’t boasting about the data during the call, though Robinson did take time to read off a whole laundry list of digital accomplishments. It only came up in response to a question about the Telegram’s performance.

But different third-party analytics companies often reach different conclusions. So if you do want to look at a different data set, here’s one from Compete, via Jonathan Mendez. As you can see, it tells a very different story–a 20 percent drop from August to September (click to enlarge):

http://www.marketingtactics.com/ davebarnes

Number of visitors is meaningless in this context.
How many PAID subscribers are there?
http://mediamemo.allthingsd.com/ PKafka

It’s not meaningless at all. If the traffic and visitor numbers hold up once the paywall goes up, then it’s very encouraging for the publisher.
http://www.facebook.com/michael.tippett Michael Tippett

Peter,

Perhaps the majority of their visitors look at less than 10 local stories a month and therefore the paywall wasn’t an issue for most of their readers. Couple that dynamic with a couple of single story hits from search engines (also sub 10 story audience) and you’ve got a formula for good traffic numbers. The question really is, did they make any money from the paywall.

Michael Tippett.
Anonymous

Um…could it be that those people who exceed the 10 free reads then begin to use a different computer or perhaps spoof their IP? Hmmm….go figure.
bridgetwi

hmmm. I am absolutely positively certain that Janet Robinson is above this of course, but some other, less scrupulous, media entity may have asked their bizdev person to strike some paid traffic deals or content barters to make sure to juice the visitor numbers because everyone would be asking if they fell. And one could ascertain this if they asked to see PV/session in launch month vs. the last. But we don’t need to do that because I’m sure that’s not the case.
http://mediamemo.allthingsd.com/ PKafka

They could! But do you think there are 13,000 readers (give or take several thousand) Telegram readers committed enough to do this but not to pay?
Anonymous

Meaningless. I bounce off of pay news sites (Wall Street Journal) all the time. I get directed there from another site. I can’t read a story. I curse. I leave.
http://mediamemo.allthingsd.com/ PKafka

Unless you have been directed to Telegram 10 times a month, you won’t be bounced off. Meanwhile the NYT model is built explicitly to let third-party referrers send the site traffic, even when readers have hit their monthly limit.
Anonymous

I think it is very easy to do and I would not underestimate the power of “free”. The bottom line is that this information tells us very little. I somehow doubt that they are getting that much legitimate traffic immediately after a pay wall (especially if that is the only thing that changed) – there are simply too many other variables that we do not know. Too many variables and too little information to say this is legitimate. What I suggested is just one possibility – there are many others (as suggested in some posts) for a spike in traffic. Nothing that would imply a pay wall is suddenly successful.
http://mediamemo.allthingsd.com/ PKafka

Fundamentally, the way to measure a paywall’s success is by looking at the paper’s financials. Unlikely that we’ll see those soon, if ever. But as a stopgap measure, site traffic is meaningful — if the wall went up, but the site is still able to attract as many visitors, and serve as many ads, as it has in the past, then that’s a very good sign.

The Telegram is 2 months into its paywall, so the real answer is that it’s way to early to draw any real conclusions about what’s happening there. But it is interesting.
http://twitter.com/joshkirschner Josh Kirschner

The Telegram analysis isn’t very useful without also understanding the percent of returning visitors, average page views per visit and percentage of search traffic. If most of their traffic is driven by search, few people would be limited by the paywall for exceeding 10 articles and traffic wouldn’t be hurt. Also would be interesting to know how many people hit that registration page when they try to access local content and go no further (yet the registration page still counts as a visit).

A Newspaper Pay Wall Goes Up–And So Do Visitor Numbers

comments so far. Add yours.

AllThingsD by Writer