In a context where paper-based revenues are generally falling, newspapers have been implementing new strategies in order to diversify their revenue sources. Digital subscriptions plays a huge role in this diversification. This is the main reason for the appearance of paywalls, in their many forms, over the past 20 years.  

These “walls” block all or part of the content on publisher’s website. Readers are then asked, generally in a rather cold and non-personalized way, to pay to gain access to the content or leave the site. And, since 2010, paywalls have been democratized all over the world. In fact, any newspaper who owns a website is needing to change their monetization method, partly as a result of the 5 years of 'all free' approach.

To understand how we got here today, we will study the history of paywalls, with a focus on the United States which represents a pioneering country for online content monetization

1997: first year of experiment

Although their implementation has been democratized in recent years, paywalls are nothing new and have emerged soon after the advent of Internet and the creation of the first American newspaper websites. Very quickly, publishers tried to implement what worked from print to digital. The first (big) daily to adopt this logic is the Wall Street Journal in 1997 with a subscription offer of $ 50 / year to have full access to their website.

Source : American Press Institute

For what results? The first year of implementation made it possible to generate 200,000 new subscriptions. 10 years after the number of subscribers on the digital exceeds 1M. While these figures are impressive, it is difficult to say whether this strategy has been a complete success for the publisher. In order to do this, it would be necessary to have a precise vision of the impact that this could have on the audience and notably on the ARPU (Average Revenue Per User).

A relatively slow adoption of paywall on the rest of the country

In 2010, only 6 of the 98 US dailies (with a daily circulation of more than 50,000 copies) had set up a paywall. This weak adoption can be explained by a (still existing) fear of observing a significant decrease in advertising revenues (due to the drop in traffic and audience) without any guarantee of increasing revenues with digital subscription. The other reason is probably that digital revenues at the time represented a small share of the overall turnover of publishers, and this was not yet a matter of capital importance.

From 2010 to 2012 we saw a strong increase in the number of paywalls, from 6 to more than 50.
This is partly due to the positive results of the dailies with paywall but also (and above all) by the greater fall in print and advertising revenues. Diversification of incomes then becomes a necessity and adoption takes place “forced march”.


2012 will be the year when the number of media with paywall exceeds those that do not have one. Digital content monetization is gradually becoming the norm (at least for publishers) until 2015 when 77 of the 98 American media had joined the rest of the pack. Step by step, publishers are testing new ideas and paywall models are diversifying.

What is the most popular type of paywall in the US?

There are 3 main types of paywall in the US (and the rest of the world)

  • The hard paywall: All the content is here reserved for subscribers, and any consultation of articles requires a subscription. It is one of the first models to have been tested in the United States, but quickly abandoned. It now represents 3% of the paywalls in the United States.
  • The freemium model: This model consists of giving free access to a part of the site and keep articles with higher added value for subscribers only (generally representing 10% to 20% of the visible articles). This is the most widespread model for the time being in France in particular, although only a few are deployed (for now) in the United States with only 12% of the publishers who put it in place.
  • The metered model: it looks like in app game system of monetization (like Candy Crush), where you can consume for free, up to a certain limit. The number of lives is here replaced by the number of articles. This is usually 10 free articles then you need to take an online subscription offer to continue. It is also the most widespread model in the United States: 63% of the media have put it in place. This is the New York Times model.

“Metered” and “freemium” models are generally the most widespread because, unlike “hard” models, they theoretically increase the number of subscribers without greatly impacting publishers’ advertising revenues. They therefore present a lower risk during a test or a first installation.

The New York Times example

In 2011, the NYT successfully introduced a metered model. At the beginning, readers was granted to 20 free articles, then that number was reduced to 10 in 2012.

It is pretty common that it goes up to 5 articles before you are gently asked to create an account to get 5 more articles. Les Echos (in France) use that system. (See the A/B test set up by the NYT in the article “The New York Times paywall“). This logic is driven by the desire to collect logged data to collect behavioral information and propose more targeted advertisements.

Source : Les Echos
Source : Les Echos

As you can see from the chart above, the model is a success and allows to maintain very comfortable incomes. Subscriptions still represent a fairly low percentage of revenues, but offset the rather sharp decline in paper advertising and a slight decrease in paper sales.

With a daily paper circulation of 625,000 copies, and more than one million digital subscribers in 2015, the NYT even has slightly more paying readers than in 1995 when 1.5M Americans bought the newspaper daily.

I reassure you, the New York Times did not find the right formula the first time. Like others, the group suffered some disappointments before finding the model that works. The NYT had set up in 2005 “TimesSelect”, which gathered all its archives and best articles behind a paywall at $ 50 / year.
This initiative was abandoned in 2007 following numerous internal discussions. The revenues generated were lower than those potentially generated by advertising. In addition, journalists were not convinced that their best articles should be read only by a small part of the audience.

Paywalls are a success! Really?

In less than 5 years, the New York Times has managed to generate close to one million digital subscribers. It’s a huge success. However, this represents only 1.6% of the audience, estimated at 65.8 million unique visitors in 2015. So there is a large part of readers who are not or little monetized when facing a paywall.

The New York Times, like the Wall Street Journal and other very big newspapers, shows better overall statistics than other titles in France and abroad. This is due to their size, the international audience they reach, and the acknowledged quality of their content.

These newspapers are therefore part of the rare media with slightly more than 1% of the audience that subscribes. The figures that we observed during our study, in France and abroad, show a very large majority of subscribers <1%. Either more than 99% of the audience does not pay to access blocked content.

Several questions remain open:

  • How to monetize the 99% of Internet users who are not ready to pay today to access premium content?
  • How can we engage them without frustrating this audience?
  • How do you turn online users into subscribers?

Paywalls in France and abroad generally lack finesse and often frustrate their audience. They often offer the same experience to all types of Internet users regardless of their affinity with the brand, their level of commitment or their intention to subscribe.

Customizing the paywall according to the level of reader engagement may be an attractive option for publishers to improve the Average Revenue Per User (ARPU). By customizing the options available to readers to access content, and by sending them the right action at the time, publishers could lift frustration while improving engagement and monetization.


  • 78% of the US media have set up a paywall, and of these, 91% have done so since 2010.
  • The most common paywall model is the metered paywall, present in more than 63% of cases.
  • The New York Times is one of the precursors on implementing a metered model.
  • Customizing the paywall according to the reader may be an option to significantly increase the ARPU of publishers

Sources :