Mobile navigation


Paywalls: why your stop rate matters

For publishers experimenting with metered paywalls, where to set the free articles limit is a key consideration, writes John Barnes. You can guess at it or you can use data, specifically the ‘stop rate’ metric, to help you make a more informed decision.

By John Barnes

Paywalls: why your stop rate matters

At the start of the last decade, in 2010, News International, and The Times in particular, was dominating media news with the anticipation of its impending paywall. The arrival of the paywall for The Times and Sunday Times online editions was viewed at the time as a media landmark. And it was believed that if the venture succeeded, other newspapers and online content providers would follow suit, and that the era of free access would ultimately be seen as a quaint anomaly.

But this wasn’t the first time that the News International group had tried to charge for an online version of a UK newspaper. In 1997, The Wall Street Journal was the first online media site to implement a paywall and News International followed in the same year, with the then owned Times Educational Supplement (TES), which worked and proved to be a success. Users paid for access and used the service to not only access content but share ideas, communicate with each other and get advice from experts.

Despite this unexpected success, the thinking at the time was that free access was the way forward, and News International decided to reverse the decision to charge and started refunding puzzled subscribers who wanted to pay for the service, fearing that without their subscription money, the service would be less well funded and consequently be less useful to them and their community.

This ‘free’ philosophy was what was fuelling the internet boom, as investors pumped money into media businesses on the overarching premise that getting more users than competitors was the only game in town and that quantity, not quality, would determine the winners in this new landscape.

This was the backdrop to the discussion behind the Times paywall with many in the media industry at the time believing that it would be impossible to persuade users who had become accustomed to free access to start paying, and that advertising would be the way media businesses would fund themselves in the future.

Looking around the digital media landscape nearly ten years later in 2019 reveals a startling change with many sites and services that used to be free, now restricting access to content and services behind a myriad of walled models.

This change is in part organic, as media owners need to replace declining advertising revenues with new sources of income, but it is also driven by necessity. While many readers tend to be sceptical of having to pay for content that used to be free, it is important to recognise that for online journalism to be sustainable, there has to be some kind of payment for content.

In the era of fake news, this realisation has been a key contributor to the change in user behaviour and attitude towards paywalls and the funding of independent journalism, as users start to realise that independently produced journalism, regardless of its generality or specialisation, is important to them.

For online journalism to be sustainable, there has to be some kind of payment for content.

Hard vs soft

A paywall requires a website visitor to pay before content is revealed. Such paywalls can either be ‘hard’ – meaning no content is available until the visitor buys a subscription, or ‘soft’ meaning some form of metering is used to restrict access, but still allow users to access content without paying.

Many believe that hard paywalls only work for brands associated with significant added value, or a specialisation (such as business news), or one that dominates its own market. They are also typically popular with titles that have had a long term paid content strategy and work in markets with a higher average subscription yield such as B2B or business information services, where the criticality of the content is seen as a significant reason to pay for access.

With a metered paywall, a visitor will be allowed access to a fixed number of articles each month after which they must buy a subscription. Over the last three to five years, metering has become more popular, as GDPR has become more important and as publishers try to counter the impact of programmatic revenue decline by collecting more first party data to improve the commercial story for advertisers. There has also been a big swing back to contextual advertising, and in the case of B2B in particular, a renewed understanding of high quality ‘known’ audiences by ad agencies and marketers.

Equally from the media owner’s point of view, content is critical to an effective online business model, because the creation of content is key to building audience, which in turn creates data that can be collected about the users, their consumption habits, interests and transactions. This data can be used to better understand the audience which in turn helps develop the content.

What many are now asking is, how to optimise and tune a metered strategy to maximise sign ups. This is a particular issue for media owners who worry that the implementation of a metering strategy will impact ad inventory and further damage advertising revenues.

This need to optimise has made the ‘stop rate’ a key metric that needs to be measured and constantly improved if metering is going to work effectively.

The ‘stop rate’ is a key metric that needs to be measured and constantly improved if metering is going to work effectively.

Know your visitors

In order to optimise the metering strategy, the first step is to honestly evaluate site traffic and determine how many of the weekly users are really the readers the brand wants. Well-optimised sites inevitably attract non-core users, as they will appear higher in search results. If the site is focused on advertising revenue, it is also likely that the editorial output will, over time, have been skewed to deliver page views.

Digital advertising is driven by page views, whereas paid subscriptions are driven by engagement (a subject I wrote about in the July/August 2019 issue of InPublishing: ‘Increasing engagement: dos & don’ts’). Publishers that are reliant on advertising will have business goals that are much less aligned with editorial priorities, whereas those that are focused on subscriptions will conversely be more focused on engagement and will have business goals that are more aligned with editorial.

Different media organisations can decide what an engaged reader is, but a simple exercise is to look at visits or sessions over a three month period and look at frequency of visit, using the GA definitions of once only, 2-3, 4-8 etc, and also through which channels, such as organic search, direct, email etc. Single visits from organic search over a three month period are disengaged and 9-14 visits from email are engaged, as an example.

This is a useful exercise as total visitors will not typically be the total opportunity for conversion, because not all of them will be engaged and not all of them will be your target audience. This information can then be translated onto a funnel of: total market, total digital audience, regular readers, stopped users, subscribers and engaged subscribers. The first two are the reach of your brand or site, the middle two are the engaged users, stopped users to subscribers are the converted users and subscribers and engaged subscribers are the retention targets.

A key finding of the report is that publishers are too generous with the free article limit.

Setting the free limit

The objective is to not only be realistic about the actual size of your real engaged user base, but also to then look at an often overlooked key metric, the ‘stop rate’.

The stop rate is the percentage of all digital users who are “stopped” by a subscription prompt, a wall, or a meter limit. The stop rate is calculated by the number of users stopped by a meter or paywall in a given month over the number of unique visitors during that period and is a key metric for calculating potential conversion rates.

In a metered model, a low stop rate is typically the result of the free article limit being set too high, which means the average user will never encounter the meter; ie. they will not be stopped.

In a whitepaper published in August 2019, the Harvard Shorenstein Centre and the Lenfest Institute explores the relationship between stop rates and article limits and how this impacts conversion.

A key finding of the report is that publishers are too generous with the free article limit (the number of articles a user can access in metered models before they encounter a wall or are stopped).

Even if a publisher is successful in converting registrants to subscribers, they are underperforming and are likely having limited opportunity to grow subscription revenue if they are failing to stop a significant number of their overall engaged readership.

The report found that of the publishers studied in a thirty day period, 68% of readers only view one article, 23% viewed between two and five articles and only 9% of users were ‘regular readers’ who viewed five or more articles in the period.

This finding is key to setting meter rates. The report shows that meter limits have been declining since 2012, when the average limit was thirteen articles. Today, the average limit is five articles, with over 57% of the publishers studied having a limit of five articles or less. This is best illustrated by The New York Times. When it first introduced its meter, users could access twenty stories for free in a thirty day period. Over time, they have progressively reduced this, to its current level of five.

Publishers should aim to reach a minimum of 5-10% of their readers to be successful with a metered paywall model. The exact number of free stories then will depend on the reading behaviour of each publisher’s audience and the channel used.

Taking 5-10% as the ideal stop rate, publishers should benchmark their own stop rates against this.

There are two basic ways to increase stop rates, either increase the engagement of users, so they are more likely to encounter the wall or reduce the number of free articles.

Various meter rules can be tested by varying the duration of the meter, to be shorter or longer than the typical thirty day period. Testing increments of 15, 20, 25, 30, 35, 40 and 45 days is one way to do this in conjunction with more detailed analysis of GA data, as site users maybe more or less frequent depending on the market, the original frequency of publication if there is a legacy print product, or the sophistication of email alert services.

The simple rule of thumb is that if a site’s meter rules are within the typical range of 5-10 articles but the stop rate is low (ie. below 5%), user engagement is the issue, whereas if 5-10% of visitors are viewing 5-10 articles and they are not encountering a meter, the access rules are not tight enough.

If a site is also using personas or cohort groups to segment audiences, different stop rates can also be tested for different user types using a dynamic wall. With a dynamic paywall, the approach for each individual reader is personalised, based on factors such as reading behaviour, persona groups, frequency of visit and channel preference amongst others. Dynamic paywalls provide a better user experience for readers as messaging is targeted and personalised so conversion rates are typically better.

Media markets vary considerably and as such, there are no perfect meter limits, but it is clear from the report that publishers with tighter meter limits will have a higher stop rate and consequently will have more potential subscription targets in the funnel, which is key to driving a sustainable subscription business.

With a dynamic paywall, the approach for each individual reader is personalised.

John is hosting a roundtable on ‘Registration walls and metering’ at the Making Publishing Pay conference at Excel London, 25-26 February.

This article was first published in InPublishing magazine. If you would like to be added to the free mailing list, please register here.