Last year, the Digital News Report by Reuters Institute showed that approximately 15% of non-paying news consumers would choose ad-free subscriptions as a reason to subscribe and pay for online news. A Swedish study from 2020 also showed 80% of people in general believe ads are irritating and annoying, and 70% said they were willing to pay an additional fee to skip ads.

These numbers made our team at Dagens Nyheter interested in launching an ad-free version of our site and app. From idea to launch, it took four months. Since November 2020, the ad-free package is a part of Dagens Nyheter’s product offering.

Apart from the technical solutions on the site and app that were needed to create an ad-free user experience, we also had to calculate the price level and potential risk of making it a loss-making business.

We started by aggregating ad data from Google Ad Manager (GAM) on the cookie level and then on the subscriber level during a month-long period. We estimated and calculated the value of a unique cookie and pageview. There were also some overhead revenues that could be split between all users.

Each subscriber had an estimated ad revenue per month tied to them, and they were grouped into percentiles according to the average ad revenue per subscriber. This showed that the top 1% most valuable users stood for 13% of all ad revenues while the least valuable half only contributed with a few percentages of the total ad revenues.

It was obvious there are large differences between how active subscribers were and how much they contributed to the total revenue. This made us aware that depending on what price we chose and from which percentile the subscribers belonged, our potential uplift revenue could vary a lot. The more valuable subscribers who subscribe to the ad-free package, the lower our margins will be. The margins were seen as the difference between the price of the package and how much the alternative revenue from ads would have been otherwise.

At the same time, we assumed the subscribers who generated the most ad views would also be the ones most likely to sign up for the ad-free offer, which became a risk we had to consider.

Pricing of the package was therefore one of the trickiest parts of the project. We wanted it to both be an attractive offer for many people but we also wanted to make sure the price was high enough to not make it a loss business. We used the Van Westendorp pricing model on the result to calculate the price sensitivity and to find the range of acceptable prices. The product was launched for an additional cost of €10 (approximately US$12).

We made a Google Colab notebook for follow-up purposes. We wanted an easy way to follow the ad-free customers’ behaviour and calculate our alternative revenue (i.e. what they would have generated in ad revenues if they didn’t pay for an ad-free subscription). So far, the difference between what the subscribers pay and what they otherwise would have generated in ad revenues has been positive.

In the longer run, I am looking forward to see whether an ad-free subscription impacts the retention numbers as well as the lifetime value. To answer those questions, we would need more long-term data though.

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .