Overview of this campaign

In early 2024, SvD assembled a cross-functional team to reimagine the way we approach paywalls. By bringing together product, UX, tech, business and analytics experts, we sought to adopt a more disciplined and structured test approach for optimizing our paywall performance.

The objective was straightforward yet ambitious: grow both our subscriber base and revenue by unlocking the full potential of our paywall strategy.

Paywall optimization is complex, with numerous variables impacting performance — from visual design, copy, and placement to the type of offer and, of course most critically, the content behind the paywall. To understand these dynamics, we systematically broke down these variables into categories and tested them one by one.

Early results exceeded expectations. By simplifying both the offer and design, we achieved a 24% increase in conversions. Testing longer subscription periods yielded even more positive outcomes, with subscriber growth in the test group increasing 400% compared to the baseline, over a time period of 12 months. Revenue metrics also saw significant improvement as experiments with subscription tiers drove meaningful increases in revenue of 8%.

After some months we saw that we had several paywall variants that were clear winners, but that we also had more parameters than we could fit into one paywall.

At that point a new objective started to become clear to us, we wanted to build a dynamic logic that showed the right paywall to the right user at the right time, with the hypothesis that we can significantly improve our subscription growth. 

Hence we set a new objective: explore a dynamic logic to leverage the full potential of our paywall variants. 

To achieve a dynamic setup, we identified what data points that were available for us to use, a machine learning model that was suitable and how to build the technical setup so that the paywall can shift depending on the data model’s assessment. To evaluate impact we also have a group that gets a randomized selection of our paywalls as our baseline.

Results for this campaign

In our first iteration we are using a bandit type machine learning model, combined with a set of data points. Our first test gave no improvement compared to our baseline, the randomized group. In our second iteration we actually removed some data points, and after this we are starting to see results. The dynamic test group varies on a daily basis from the same conversion as the test group, to as much as +15% better. Our next goal is of course for the dynamic model to always be higher. 

During the coming year we see a lot of exciting opportunities, to both add more relevant data points to the model, but also to let the model reward total customer lifetime value instead of conversions. We think we have created a structure that gives us a lot of room to test, and that the test speed can be fast. 

To summarize the initiative has delivered important insights and results that lay the groundwork for sustained growth at Svenska Dagbladet

1. A sustainable method for testing: We have achieved a way to continuously iterate and learn to maximize the potential of our paywall 

2. Original objective, identifying new paywall variants: We created multiple highly effective paywall designs tailored to our business goals, increasing conversions, revenue and subscriber base

3. New objective, a dynamic paywall logic: By exploring technical setups and machine learning models, we implemented a solution capable of dynamically adapting paywalls to user behavior, driving additional growth

4. Strategic long-term potential: Early tests show opportunities for further iteration, with strong hypotheses on how to improve

These advancements have not only delivered strong results in 2024, but have also provided us with a scalable framework for future innovation. With these developments, Svenska Dagbladet is well-positioned to achieve its 2025 business targets, demonstrating how a strong paywall strategy can play an important role in securing long-term financial results.

 

Contact

To contact a company representative about this campaign, click here for the INMA Member Directory