Twitter  "Algorithms have consequences." Zeynep Tufekci on Ferguson and net neutrality:  
Nieman Journalism Lab
Pushing to the future of journalism — A project of the Nieman Foundation at Harvard

So your newspaper is thinking about putting up a metered paywall. What kind of results can you expect?

The answer will depend on a lot of factors — the makeup of your audience, how tight or loose you set the meter, what you’re charging. But from looking at other newspapers’ results — more than 500 have a paywall in the U.S. alone — you can start to estimate.

Here’s one data point, from paywall service provider Press+, which runs hundreds of those sites. Of a site’s monthly uniques, about 3 to 4 percent will hit the paywall — that is, they’ll consume enough content to be told they have to pay to read more. And about 0.5 percent of your monthly uniques will hit the wall and actually pay.

That’s from this article by Jeff Hartley, vice president of consumer revenue for Morris Publishing Group, at the Newspaper Association of America site.

With the help of our digital subscription management partner, Press+, we know what percentage of our unique visitors is encountering a stop light box (SR) and how many, once stopped, are converting to a subscription (PSCR). Because Press+ has so many affiliates, 450 and counting, they’ve established performance benchmarks for both the SR and the PSCR. These indicators are crucial to knowing how effective you are in engaging readers and converting them to paid sales (Note: these metrics can be set these up on your own using Google Analytics, Omniture, etc.)

Based on data from Press+, the average publication has a stop rate (the number of unique visitors that see a stop light box) of 3%-4%, but high performers stop between 5%-10% of unique visitors. A PSCR rate (the number of unique visitors that are stopped and then purchase) of 0.5% is considered average, while high performers convert 1%-2% of stops into paying subscribers.

(You may notice that that last sentence is a little screwy — the first half says it’s measuring the share of unique visitors, while the second half measures share of stops. I’m pretty sure the second half of that last sentence is misworded — it should be uniques. Otherwise we’d be talking truly tiny numbers — 0.03 × 0.005 — which might buy you a couple cups of coffee, but not much more.)

So, back of the envelope math: Your site gets 1 million monthly unique visitors. A typical metered paywall would be invisible to the vast majority of them — they won’t read enough stories to hit it. On average, you might expect 30,000 to 40,000 to hit the paywall — but that could be as high as 50,000 to 100,000. Of those stopped, you’ll probably get 5,000 to pay up — if you’re a top performer, maybe 10,000 or 20,000.

Still, even within Press+ customers, there’s a wide range of outcomes; as Ken Doctor reported for us last November, the best Press+ paywalls far outperform the worst. The details matter:

While each publisher gets its own data, Press+ gets it all. Out of that, you get one of the major selling points Press+ uses with its clients: “We’ll share best practices with you.” Press+ has measured the differential between its highest- and lowest-performing sites at 10-to-1, and that would be an impressive difference-maker in revenue return. Get the metering, marketing, messaging, and more right, Press+ says, and you’ll more than make up the revenue share (~20 percent) you pay us.

— Justin Ellis
What to read next
BuzzFeed HQ
Caroline O'Donovan    Aug. 11, 2014
With a new round of investment from Andreessen Horowitz, the viral giant aims to get even bigger. One step down that path: making it more clear to readers what kinds of content BuzzFeed really offers.
  • Seb Chan

    Going from ecommerce data I’d think the conversion rate of those who actually hit the paywall is not incorrectly expressed – and it *is* minuscule and thus relies on sites getting a huge amount of traffic. Can you get clarification from Press+ because it is key point?

  • Informerly

    Agreed if thinking about ecommerce conversion rates. That 1mm uniques, on the high end would maybe result in 1-2k subs (100,000 stops, 1%-2% conversion) sounds right.

    In the premium subscription news world, even with all qualified leads in the funnel, conversion rates around 10% were incredible. And that was leads that had been pre-screened, not completely random uniques.

  • mattdelman

    The wording makes similar sense from an email marketing perspective, so I’m with you on it not being incorrect.

    I got a base PSCR rate of 150 a month at the truly low end, which doesn’t sound too out of the realm of possibility.

  • Joshua Benton

    If it really was 0.5% of stops, that would be around 150 paying subscribers per million uniques. I can tell you that, as low as newspaper digital sub numbers can be, they’re a lot better than that. Many metro newspapers are in the tens of thousands of digital subscribers; The New York Times is over 700,000. Their monthly uniques are nowhere near high enough to make the numbers work with that math.

  • Bharat Ayyar

    You’re right. The conversion rate as Jeff expressed it is correct. Our leading sites stop 5-10% of their unique visitors. And 1-2% of those convert to paid, digital-only subscribers. (The conversion rates are higher if we include print-and-digital bundles — but here, Jeff was just referring to digital-only subscriptions.)

    Another misinterpretation is that this is NOT a lifetime sales projection, but rather a monthly conversion rate–a measure of sales and marketing effectiveness. In other words, if you are a leading Press+ performer with 1M pageviews, you will stop 50,000 unique visitors (5%), and convert 1,000 to paid subscribers (2%) EVERY month. Again, we’re talking just digital-only subscribers here. Almost all of our publications sell bundled subscriptions as well.

    The conclusion is roughly the same: Over the course of 2 years, you will have converted roughly 2.5% of your total unique visitors to digital-only subscribers.

    Bharat Ayyar
    Analytics at Press+

  • Bharat Ayyar

    The statistic quoted is a MONTHLY metric not a lifetime metric. Please see my other comment.

  • mterenzio

    That’s for a leading performer, but the average on those numbers is actually around 250 subscribers a month , with many doing much less.

  • The Spain Report

    Hi Bharat, could you perhaps shed some light on what it is the top performers do differently than the rest, please? A Top 3 or 5 list? E-mail list? Site design? Offers? More aggressive marketing? More close and tribal? More frequent article publishing? How are they getting those better numbers? Thank you kindly.

  • Laurent SICSIC

    I agree. The conversion rate expressed by Jeff Hartley seems correct to me. In fact it is roughly the metrics we get in Europe with digital newspapers/magazines.

  • José Blanco


    Please, could you explain the measurement methodology for the “total unique visitors” .
    What traffic is included ?


  • Curtis

    Hi Bharat, my question is in regards to the definition of unique visitors you are calculating off of. Are you using all unique visitors, including subscribers and “fly-bys” (those who visit just 1 time per month and typically make up a large percentage of total unique visitors) or just the non-subs and non-fly by unique visitors?

  • Anne-Marie

    Hi Bharat – I have the same question as Curtis – how are you defining unique visitors for your calculations?

  • Dan Gayle

    This is great, except for the fact that this almost reads like an ad for Press+.