SME (Dennik SME) is a subsidiary of Petit Press and the second most read newspaper in Slovakia. Its digital arm, SME.sk, is the most widely read digital news portal in the country. This case sutdy is based on the presentation at INMA’s Media Subscription Week 3.0 in NYC.

SME.sk utilizes a distribution strategy offering ad-supported free content as well as subscription-only premium content. Under this model, a user can read an unlimited amount of free articles on the website. However, when they click on any premium content links, they will be offered a hard paywall. If readers refuse to subscribe, they may simply return to the free content and continue reading. SME used their CMS paywall for premium articles and Piano for the paywall displays on free content.

What was the challenge?

• SME manages a standard freemium paywall with:

• 80% of content free to all users and supported by ads
• 20% of content hard locked behind a premium wall – articles are chosen by the editorial team

• It is a not uncommon problem with freemium models that highly engaged users with a great frequency of visits simply never move up the funnel and subscribe.  

• However, considering their popularity with users was there a way for SME.sk to strengthen their subscriber base?

Could SME acquire more subscribers by limiting access to free content for the most engaged users?

How was this project implemented?

SME implemented a hybrid premium / dynamic paywall that:

1. Retained premium content - 20% of all content, chosen by editorial, was locked for all users
2. Dynamically locked non-premium content for users who had a high Propensity to Subscribe (P2S) only on content that was on the subscription path of many users
3. Kept remaining content - non-premium content not likely to convert - open to all users
4. The goal of the implementation was to test a hybrid (premium + dynamic) paywall that would increase conversion and boost subscription rates.

In basic terms, the trial would dynamically close content based on the user’s Propensity to Subscribe (P2S) score.

Hybrid Premium/Dynamic Paywall – Project Setup

The Propensity to Subscribe (P2S) Algorithm: Deep.BI provided real-time user scoring that was integrated with SME’s paywall

Deep.BI created a Propensity to Subscribe score; a machine learning algorithm built on 100+ features, where user engagement, loyalty, and habit have the greatest weight.

A/B testing: Premium content (20% of all content) remained hard locked behind the premium paywall. Non-premium content (80% of all content) was either open or closed based on Deep.BI’s P2S-driven Dynamic Paywall. The Dynamic Paywall actively locked non-premium content for users who had a high Propensity to Subscribe (P2S) score and whose reading journey followed the subscription path of many other users

For every website or app action a user takes, Propensity Scores are immediately recalculated and pushed to the paywall for automated, real-time decision making

The Deep.BI Hybrid/Dynamic paywall was shown only on free content when users reached a highly engaged level. Alternatively, different scores are calculated and pushed to different systems (CRM, Email, ….)

Find out more about our propensity scores

Hybrid Premium/Dynamic Paywall – Project Results

Results: Overall

The Deep.BI Hybrid Premium Dynamic Paywall:

• Was shown 20x less often than the SME premium one, and it targeted only highly engaged users
• Was 278% more likely to be clicked (CTR)

• 50% more users finished the purchase process via this route compared to the premium paywall
• 66% of users who purchased a subscription from premium articles were successfully predicted to do so by the P2S algorithm

There was a 50% higher purchase completion rate from users who saw the Hybrid/ Dynamic Paywall and a +200-300% paywall CTR.

Results: User Engagement:

• No overall observed decrease in page views
• Average User Engagement scores remained the same
• Both metric values increased by 10-15% because content access was not limited by the paywall
• Deep.BI’s propensity to churn model predicted a 12% lower churn rate for those who subscribed because of initial dynamic targeting

SME saw a +15% increase in users in the engaged segment.

Results: Display advertising unaffected

• During the test, the average number of page views per user was identical, so the dynamic paywall did not decrease display advertising yield
• Users impacted by the dynamic paywall generated the same or greater number of page views and visits

There was no negative impact on advertising during the testing period.

Other benefits for the client from this type of solution?

• With the ability to utilise a mixed-paywall solution, SME will be able to maintain Premium Content Protection from users in Incognito mode aka ‘Clear Cookie’ anonymous traffic.
• Additionally, SME retains editorial control over its Premium content with an option to exclude exclusive content from dynamic mode and establish a hard paywall rule for chosen content.

Implement a dynamic paywall today

Deep.BI provides publishers real-time user scoring Powering Subscriber Acquisition & Retention. Using our Machine Learning algorithms, we build Propensity Scores for Subscription, Churn, and even Registration to bring added value to your user journey across the board. Connect Deep.BI to your tech stack and immediately start scoring your users and automating actions in your Paywall, Marketing Automation systems and more!

‍