For several years, publishers have set up paywalls to develop digital subscriptions and diversify their sources of revenue (see the state of paywalls adoption in the United States). This is increasingly present and scrutinized in the media industry: how to find the right formula to transform readers into subscribers?

We experienced an intense first period during which the (online) paying model appeared, intensified, present in different way (freemium, metered, hard, datawall, …). Although present in different models, paywalls remained “simple” because offering the same experience to all readers. A “one fits for all” model that allowed publishers to frustrate 1–2% of their core readers and turn them into subscribers. Today, subscriptions represent a huge potential for publishers, and probably a healthier model where readers pay for content.

A legitimate question arises: how can we go further to develop digital subscription?

To illustrate this question, I personally like Frederic Filloux sentence: “let’s fill the rest of the plane” in his same name article. Meaning: we sold a good part of 1st class seats, how to make money with the rest of the plane?

Use data to drive paywall

Publishers use different tools to collect, analyze and use their 1st party data. The data is, for some years, used on the advertising side to better value inventory and fetch better CPM.

If we go back to our paywall story, the “one fits for all” model works well for the fans but clearly represents a friction zone between readers and publishers for the rest of the readers. Early studies with our customers show that, on average, 60% of readers leave the site directly when facing paid content (which has an impact on the publisher’s entire business model).

And it’s quite logical when you look at how a publisher’s audience is composed. It’s finally broken up into ten smaller user segments. Old subscribers, fly-by readers, those with an adblock, fans, logged ones… All with an specific history, a certain relationship, a different affinity with the brand. Different levels of commitment and probably a different willingness to subscribe. Paywalls today have a very limited and inadequate response for all of these readers.


David Skok, former head of strategy at the Toronto Star, sums up the situation very well in a tribune on the Nieman Lab site : “All of it makes the notion of having binary on-or-off payrolls and press releases touting.” 10 free articles a month “seem antiquated.”

We must build the paywall of tomorrow that will boost publishers’ revenues. Customizing the paywall can and should allow publishers to transform the friction zone that the paywall can represent into a dialog box. While keeping it simple.

Now come the question: “Okay, all of this is very interesting, but concretely what should we do? How to customize our paywall and improve our engagement and revenue KPIs?”

Create custom journeys and personalized marketing offers

A good way to try to convert more readers into subscribers is to offer different offers and subscription messages depending on readers’ profile. That’s what the Wall Street Journal did recently.

The WSJ has set up a mechanism where each subscriber can share an article for free to another non-subscriber. When the non-subscriber reads the article, he can read it for free. If he wants to read a new article, he will find a hard paywall. But unlike a classic reader who is offered the classic subscription offer, this reader is offered a “guest pass” for 24H. The WSJ observed that this guest pass converted 6 times more than the classic offer.

We can easily imagine other journeys depending on readers profile :
– Suggest to readers with adblock to subscribe for an ad-free offer
– Offer to fly-by readers, occasional readers or even social readers a guest pass to discover the offer
– Offer to a former subscriber a “try us again” offer with a 1€ offer for 3 months
– Trigger the paywall (on a metered model) with a different number of article depending on readers’ intention to buy

To check the impact of a message or a subscription offer, it’s strongly recommended to A/B test different options on these different segments to ensure the relevance of an offer, a message or not. And to stay simple?

Offer customized pricing to access content

If we look to the ​​”personalized offers” concept a little further, we can easily imagine offering different prices, according to a multitude of criteria. A sort of “dynamic pricing system” such as airlines yield management. The one who talks about it the best remains Frédéric Filloux in his article (which I strongly invite you to read) “How Facebook and Google could disrupt the subscription model for news“. He takes as an example Christian Popp who is the director of revenue and strategy within Lufthansa: “His tips focused on business practices that we have in common and was surprisingly relevant: Get your customer segmentation right and you‘ ll grow your revenues if you run tight control of your inventory. Firstly, create an accurate segmentation of consumers and trade clients, based on their willingness to pay. Knowing how to buy and sell products and services.”

Since the additional cost of producing a digital subscription is close to zero, this method would increase the conversion to the subscription and thus the publisher’s revenue to the global, even if the unit value of the subscription would drop.

Provide readers personnalized compensations options to access content

If you analyze the propensity of your readers to pay at T time, it is likely that a large majority will not be willing to do so. That does not mean they will not do it tomorrow. In any case, it is certain that these readers, the most volatile ones, are not today monetized through paywalls. The single response model may even have a negative impact on readers’ commitment (60% leave the site directly) and their propensity to subscribe tomorrow.

David Skok illustrates clearly this reality in his speach on the Nieman lab website: “Imagine a reader browsing the web on their smartphone while on a train heading into work. They click on a link to see where they are served at paywall. Using predictive analytics, we are quite certain that this will not happen. In fact, the reader may even post on Reddit just how much she despises your paywall.”


We could offer personalized and durable compensation options other than subscription at first to monetize readers and most important, engage him gradually with the brand: “So, instead of wasting our time trying to get that reader to subscribe, what other kinds of value can you exchange with her that could be of benefit? Perhaps it’s an email newsletter signup form that could begin an inbound marketing relationship? Perhaps it’s a video ad with a high CPM to generate maximum ad revenue? Perhaps it’s a good idea to have a look at it.”

Many other scenarios can be imagined, in addition to those advanced by David Skok, such as offering readers with an adblock to whitelist publisher’s site in exchange for article access or even answering a question to collect data. Or offer a free article to a reader who discovers the website for the first time in order to start a dialogue, a relationship with him...

The result is a better experience and increased revenue per user.

A study conducted by Scout Analytics with a US publisher determined the revenues generated by different audience segments. Audience was split into 4 large families (fly by readers, occasional readers, regular readers and fans), depending on their level of engagement with the media. Obviously, the more engaged the reader is, the more money he brings to the publisher, until he subscribes (called fan here). More information on calculating ARPU can be found here.

Paywall customization, with the various examples mentioned above, has a twofold goals:
1 — Improve the users’ experience and therefore their commitment
2 — Increase revenue per user either via subscription or other publishers’ BMs

Let’s imagine a publisher who is able to ‘move’ a little less than 300 000 UV of volatile readers to occasional readers, regular readers or fans (less than 2% of the total population). When we look at the results below, we see that this has an enormous impact: a 30% increase in turnover.

Regular and casual readers should also be seen as potential future subscribers. Indeed, we observe with our customers (and very logically) a propensity of the reader to subscribe proportional to his level of commitment.

Paywalls should be a dialog box and not a frustration zone between readers and publishers. And paywall customization must be the key that allows publishers to improve the experience and develop their income more significantly.

If of course, personalization remains simple and if publishers have a tool to control the paywall with their data.