On its surface, the definition of a paywall is straightforward: A digital popup that blocks you from reading the story you clicked on unless you pay up.

But the catch-all term “paywall” conceals several smaller decisions that go into designing exactly how much a reader sees as a news org coaxes them to subscribe. Do they see an image, like I saw when I clicked this Los Angeles Times story? Nothing but a headline, like this Financial Times preview? A headline and image, like this Boston Globe story? Or the headline, image, deck, and lede combo, like this Wired story or this one from McClatchy-owned Raleigh News & Observer?¹

Which of those visible elements are most likely to yield success? That’s one question the authors of a study recently published in Journalism Studies aim to answer. The study, “Converting Online News Visitors to Subscribers: Exploring the Effectiveness of Paywall Strategies Using Behavioral Data,” looks at millions of visits to 21 German and Austrian² local and regional news sites in 2022 to gauge whether certain “teaser” features of a paywall (decks, images, ledes, a blurred preview of an article) made users more likely to click “subscribe.”

After that first click, visitors typically confront specific subscription offers. So the study also examined a second question: Which incentives — small gifts, ePaper access, and so on — made users more likely to purchase a subscription?

Their findings, in brief: Less is more, when it comes to teaser features in front of a paywall. And discounts are the most effective enticements for prospective subscribers.

“Reducing the information density of teaser elements on paywalled articles and offering discounts may help newspapers increase their online subscriber numbers,” the paper’s authors (Zhengyi Xu, Neil Thurman, Julia Berhami, Clara Strasser Ceballos, and Ole Fehling) write. They emphasize that their research captures actual user behavior, not just what people say they do.³

The study’s authors examined how each “teaser” element contributed to a visitor’s likeliness to subscribe. They found that decks (“stand-ins,” for those across the pond) and intros decreased the odds of a visitor subscribing by 86.3% and 72.2%, respectively. (Results for an image and a blurred article preview were not statistically significant.) Readers who see information-dense nuggets like ledes and decks, the authors write, may feel they’ve learned all they need to know in front of the paywall — and don’t need to subscribe to learn more:

Visitors may obtain enough information from a well-formulated stand-first and/or intro to feel that they have been sufficiently informed, and therefore feel no need to read further, let alone pay to do so. This is especially the case given online news consumers’ tendency to quickly browse news to form an impression of current events, rather than seeking in-depth knowledge or opinions (Costera Meijer and Groot Kormelink 2015).

Meanwhile, when it comes to the second step of actually purchasing a subscription, “discounts increase the odds of paying for a subscription 3.35 times.” (None of the other results for incentives tested were statistically significant.) Here’s what the authors say about those findings and potential limitations:

Offering a direct discount may provide visitors with the feeling that they are receiving a tangible benefit and encourage them to subscribe. This aligns with previous studies that show price discounts have a facilitating effect on consumer purchases of products and services, including newspaper subscriptions (Al Hafizi and Ali 2021; American Press Institute 2017). However, this finding may not generalize to markets (e.g., Poland and France) where significantly higher or lower proportions of online news subscribers receive discounts than they do in Germany and Austria (Newman et al. 2024).

Because the correlation between trials and visitors subscribing was not statistically significant, “a shortening of the trial period might generate more income for news publishers without having negative effects on the propensity to subscribe,” the authors note.

The authors separately examined local and “remote” visitor behavior. While both groups were less likely to click “subscribe” when paywalls showed a deck or lede, there was a difference between the groups in the appeal of discounts: “Regarding finally converting to a subscriber, the results show that for local visitors, none of the offer/price strategies are significantly correlated. However, for remote visitors, both a discount and the giving away of free or subsidized smart devices significantly increase the odds of visitors finally deciding to subscribe.” To explain that discrepancy, the authors suggest “local visitors may care only about the news itself and grant little weight to external factors…remote visitors may lack the same geographical connection with some of the news content on the regional and local news sites in our sample, and so may need more external factors, such as discounts and tangible benefits, to find subscription attractive.”

These findings pose a version of the dilemma that paywalls already embody for news outlets. Paywalls are a tradeoff, generating revenue while cutting off access to news for those who can’t or don’t pay. Some newsrooms hope to move away from them completely; others are embracing them (or making them smarter) and see them as an essential evolution for future sustainability. (And many drop paywalls on breaking news affecting public safety, as we’ve seen with the pandemic and, more recently, the Los Angeles wildfires.)

The authors suggest in light of their findings, “reducing the information volume of these teaser elements could also make sense for news websites. As many of the news websites that we studied showed both a stand-first and an intro on their paywall locked article pages, another possible strategy would be for the publisher to remove one of them, or even both.” If a key implication of this study is that news orgs looking to get people to subscribe should keep even context-heavy decks and intros cloistered in the walled garden of news for the best chance of enticing more subscribers, that could diminish access to free, reputable news even further.

Is a newsroom’s only goal with a paywall to turn visitors into subscribers? Or do news orgs want visitors who hit a paywall and don’t subscribe to still leave with a little more context than a headline alone can offer? That’s a question that is more about mission than data, and might be trickier to answer.

1. YMMV — you may not hit a paywall on first click, since many of these publications employ metered paywalls. [↩]
2. Germans and Austrians are slightly less likely to say they pay for news than the global average, according to the 2024 Reuters Digital News Report. “This may mean that it would be easier in some other countries to nudge users to start and finish the subscription journey than our results, based on German and Austrian users, indicate,” the authors write. [↩]
3. A bit more about the study’s data source: the authors relied on data from the Digital Revenue Initiative (DRIVE) Project. Launched in 2020, DRIVE “tracks and analyses publishers’ online news content and user behavior in order to help publishers take actions to increase digital revenues” in partnership with more than 25 publishers, according to the study authors. As of May 2022, when the first data was collected, “the project had recorded 4 million pieces of article information and 10 billion user events.” In this study, “regarding whether the ‘subscribe now’ button was clicked, 3,518,822 individual-level cases were included, and regarding whether a subscription was taken out the number was 28,506.” (Since this data used cookies to establish the identity of non-subscribing visitors, individual users are technically “individual browsing devices.”) [↩]