Articles by: "Matt Lindsay, President—Mather Economics"

Subscription motivation and acquisition tactics

February 12, 2020 Matt Lindsay, President—Mather Economics

As the art and the science of acquiring digital subscriptions has progressed, the use of advanced analytics to profile potential subscribers and target them with subscription offers has become standard practice. The early days of 20 articles-per-month metered access have given way to hybrid Freemium models that offer premium content to subscribers in addition to metered access to free content for non-subscribers. And certainly, balancing advertising and subscription revenue remains paramount to many publishers. This is particularly true for those with advanced programmatic advertising revenue capabilities.

My last post discussed the tactics for maximizing total digital revenue, audience, and advertising revenue, by using an intelligent paywall strategy. This post builds on that discussion with additional insights on how audience analytics add value to the subscription revenue model. The goal is to help publishers refine their acquisition strategies to reflect the motivations of different customer segments for subscribing. These same analytics can help publishers take steps to retain customers once they have taken the leap of subscribing.

Propensity to subscribe

Our recent work with publishers developing propensity-to-subscribe models has identified several reasons why individuals subscribe to digital content products. These factors have also been written about by the American Press Institute in their research on reader revenue that used surveys of recent subscribers.

These factors are often cited by many subscribers as their reason for subscribing:

Community – Refers to a reader’s interest in local news and events. 60% of the API survey respondents mentioned local news as a reason they subscribed.
Cost – Receiving a discounted offer at the right time is also a common factor in subscription decisions. 45% of the API survey respondents mention discounts or promotions as a reason.
Content – The desire to read content that the reader finds interesting. 40% of survey responders gave this as a subscription reason.
Contribution – Supporting local journalism, whether the reader is actively engaged with the publication or not, was mentioned by 31% of survey responses.

Key variables

In our econometric models of subscription propensity, these same reasons were identified as important for predicting customer behavior. Local news content is consistently one of the significant drivers of subscription purchases.

Offer price is also an important factor in our models of propensity to subscribe. Variables that are correlated with interest in content, such article page views, average time on site, and frequency of visits are important predictors. These statistical results developed on observed customer purchase decisions validate the stated subscription reasons in the survey data.

Harder to measure empirically is the influence of the contribution motive on an individual’s likelihood to subscribe. However, we do observe a consistent proportion of digital subscribers that have very low engagement following their subscription purchase. This behavior supports the belief there is a large audience segment that wants to support local journalism despite low engagement with the product.

Understanding value

As an interesting aside, there are precedents for people placing economic value and paying for resources that they do not use. Environmental economists measure the value individuals place on the fact that certain natural resources exist, such as the Everglades or Redwood forests, even though they may never see those resources or use them directly.

Understanding the motivating factors for subscribers is helpful for selecting acquisition and retention tactics to implement, and it helps explain why freemium models are replacing metered models at many publishers. For those readers motivated to subscribe by interesting content, premium content is an effective acquisition tactic as readers motivated to access a particular article will accept a subscription offer. Premium content also gives those wanting to contribute to the journalistic mission a low-friction path to subscribe relative to a multi-article metered strategy.

Pathways to subscription

For publishers that have had a premium content strategy, the trend is to increase the proportion of content categorized as premium. Le Monde recently shared they now have 35% of their content as premium, even though it may negatively affect advertising revenue. Schibsted in Norway has also increased the percentage of their content designated as premium.

We often find that less than half of new digital subscriptions are sold via a paywall. The other subscription sales come from newsletters, emailed promotions, and the “subscribe now” option on the site. Personalized newsletters and email promotions can focus offers on those readers who have hit the meter one or more times but not subscribed at the standard offer. Paywalls can provide this type of offer targeting, but there is a risk in having dynamic pricing on a site itself as many customers find this type of targeted acquisition pricing confusing. It is also hard to implement since many customers come to a site from different devices and browsers.

The current state of the art is deciding what content is premium to each reader at what time. There are a growing number of publishers using predictive analytics to assign content as premium based on its expected subscription conversion performance. However, most often the editorial staff are making these premium content decisions. Combining this tactic, dynamic premium content decisions, with an intelligent paywall can maximize subscription sales while still protecting advertising revenues.

My next article will describe how retention tactics that reflect these subscriber motivations can be applied to each customer segment to minimize digital subscription churn.

Advertising | Revenue

How intelligent paywalls maximize total digital revenue

October 2, 2019 Matt Lindsay, President—Mather Economics

Audience revenue is a growing component of the digital business model, but it often comes at the cost of advertising inventory. News media companies are struggling to support their newsrooms with digital advertising revenue and subscriptions offer a stable revenue source. In fact, media companies of all kids are seeking subscription revenue as they find digital advertising revenue insufficient for sustaining their operations.

However, as they employ a subscription revenue strategy, publishers must consider the potential for lost advertising revenue from reduced site traffic and address it through smarter paywalls. As digital publishers limit the public’s access to their content by putting it behind a paywall, they place some advertising inventory at risk. A membership model strategy, such as that pursued by De Correspondent in the Netherlands, or a contribution revenue strategy, pursued by The Guardian in London, may alleviate some of this lost advertising revenue. But all audience-based revenue strategies will have some effect on advertising inventory.

Advertising revenue will also be at risk due to the product changes required to make a digital product compelling to a paid audience. Improving the user experience typically requires reducing the number of advertisements on a site relative to a pure advertising-revenue business strategy.

Two different stories

A sample analysis of two news media company’s digital audience and advertising revenue is presented below to measure the potential return from employing an intelligent paywall in place of a standard paywall. The data has been adjusted to protect confidentiality, but the relative metrics across segments remains accurate.

Publication One is the second-largest newspaper in a large American city with about 10 million average monthly page views and about 4.5 million unique visitors. It is focused on long-form articles. Publication Two is a regional publication with about 120 million average monthly page views and about 22 million unique visitors. It has infinite scroll and publishes shorter articles. Neither publication had a pay wall at the time of the analysis of their audience.

The current audience for each publication was segmented by engagement. Engagement is defined using three dimensions: the volume of content consumed, the pattern of consumption behavior, and the nature of the content consumed. There are six metrics helpful for quantifying these three dimensions:

Article page views
Frequency (how many visits per month)
Recency (average time between visits)
Time on site
Content breadth (number of content categories visited)
Scroll depth

These engagement metrics are used by many publishers to segment their audience, and they are correlated with a customer’s propensity to subscribe.

**Engagement-based audience segmentation of two news media sites**

In the table, we see that engagement decreases from left to right moving from the Fanatics to the Non-engaged. Metrics for each segment are reported for a 30-day time period. Fanatics and Enthusiasts are the customers most likely to subscribe. Stable Users and Dabblers require more engagement before they are viable subscription candidates. Fly-bys and non-engaged are unlikely to be observed again beyond the one visit captured in this data. Using these metrics, it is possible to estimate a propensity to subscribe for these segments. We are also able to measure advertising revenue from a user in each segment[1].

The propensity to subscribe can be estimated using several approaches, including machine learning methods and econometric models. The segmentation is determined by relative propensity to subscribe, and differences in audiences can be seen in the metrics. For instance, the Fanatics in Publication Two have more article views than the same group in Publication One. This is in part due to the nature of the two products and the mix of mobile browser versus desktop browser traffic. Other behavior metrics, such as time on site and scroll depth, are affected by the nature of the site user experience and hard to compare across sites. However, visits and unique days are helpful for comparing the two audiences. In these two metrics, the audience for Publication Two is found to be relatively more engaged for a comparable level of subscription propensity.

Engage and convert

As expected, the audience segment most likely to subscribe are the Fanatics, which average 2.5% of the site users. The Enthusiasts are the next most likely group of subscribers, with an average of 4.5% of the users. Together, these two customer segments represent the majority of the addressable market for acquiring digital subscribers.

Stable users are potential subscribers. But the best tactic for that group is to grow their engagement via registration, newsletter promotion, and content recommendations before seeking a subscription. Publishers that have print products and all-access subscriptions may have a significant portion of these addressable customer segments already as digital subscribers. Some of these customers may be economically unable to subscribe, and many digital readers will take a long time to convert to subscribers or will not subscribe due to the value proposition of the current digital product.

An important observation from these data is that the customers most likely to subscribe generate a significant share of a publication’s advertising revenue. Asking for subscriptions from this group would appear to significantly reduce advertising revenue. However, not all of the ad revenue from these customers is at risk from a metered or premium content model. Pages that will not be behind the pay wall, such as the home page and index pages, make up a large share of the ad revenue. Other types of ad revenue not affected by a pay wall are also removed from the at-risk revenue. The chart below shows how the at-risk advertising revenue is determined for a site.

**At-risk advertising revenue from a pay wall**

The target groups for subscriptions will have relative high conversion rates, which will also mitigate the loss of advertising revenue. When the at-risk advertising revenue is calculated for the segments and adjusted for likely conversion rates, we find that the revenue at-risk by segment are about equal, as is observed the following chart.

**Advertising risk by engagement segment**

Balancing risk and rewards

A conclusion from this analysis is that an across-the-board pay wall will increase the advertising revenue loss. If the pay wall limits access to only those customers likely to subscribe, the lost advertising risk will be significantly reduced. Another conclusion is that the across-the-board pay wall will attempt too few subscription sales to the target audience because it is balancing the revenue streams across all customer segments and not just the segments likely to subscribe.

To illustrate these points, the table below gives forecasted results for two alternatives for a site currently using a 10-article pay wall. The first alternative reduced the number of articles a reader will receive before reaching a pay wall modal with subscription offers from 10 articles to 5 for all users. The second is a targeted meter of 5 articles to only those customers in the likely-to-subscribe segments and a meter at 10 for all other readers.

**Comparing a targeted pay wall strategy to the current state and across-the-board meter of five article views per month**

The key findings from these alternative pay wall scenarios is that targeting the readers most likely to subscribe with acquisition offers increases the number of acquired subscribers by 11% while saving about 30% of the lost advertising revenue relative to the more aggressive across-the-board strategy and raising the net digital revenue by 27%.

A metric for measuring the effectiveness of a subscription acquisition strategy is the ratio of incremental subscription revenue to lost advertising revenue. In the table above, this metric is labeled the Dollar Tradeoff, and the use of targeted acquisition offers yields $5.98 of subscription revenue for every dollar of advertising revenue lost compared to $3.78 for the more aggressive across-the-board strategy.

Devising a strategy to maximize audience revenue while minimizing lost advertising revenue is the challenge facing digital publishers. Finding the best path forward requires innovation in product design, content marketing, and subscription sales tactics. An often-overlooked aspect of a successful digital subscription model is retaining customers once you have them. Understanding the economics and engagement of the current readership is the starting point of a digital audience revenue strategy, and an intelligent paywall is an important tool for implementing that strategy.

About the author

Matt Lindsay is the Founder and President of Mather Economics, a consulting firm that works with publishers around the world on audience analytics and subscription revenue strategies. Its clients include the many of the largest news media publishers in North America, South America, Europe, Asia, and Australia/New Zealand.

[1]
This data comes from the Listener data capture tagging solution. Ad impressions are captured in the data with JavaScript tags and revenue per ad is added to the data from the advertising billing systems, advertising platforms, and other sources.