3 Benchmark Insights on Reach and Subscription Growth

It’s planning season right now, and naturally the question arises: “Where do we currently stand, and what should we tackle next?” Besides potential AI projects, questions about audience growth and subscription acquisition are likely at the top of the list.
Our benchmark of local and regional publishers provides some insights. The analysis for September highlights three key points, which my colleague Joachim Dreykluft presented in detail last week during our Upscore webinar.

First - The share of Google Discover traffic for local and regional publishers has recently declined. This is surprising because, unlike Google’s AI Overviews and AI Mode, no major changes to Google Discover in Europe are known. Therefore, we can only work with a hypothesis: since March 2025 (when the first reports about Google Search’s AI overhaul appeared), editorial teams have focused heavily on Discover. This effort clearly paid off during a very successful early summer. It seems that some editorial teams have since reduced their Discover efforts, which is immediately reflected in the numbers. One indication is that some individual offerings dropped more sharply than others. In our view, Discover should continue to be pushed with full force.

Second - Relative reach (i.e., the number of article views per inhabitant in the distribution area) has continued to improve. The median is now 6.5. According to our analysis, this is generally sufficient to generate enough subscriptions. Our recommendation: don’t focus excessively on building additional reach alone, but simultaneously focus on subscription acquisition.

Finally - The data clearly show a weak correlation between reading time/media time and subscription conversions. In contrast, the correlation between “content gating” (Verplussung) and subscription success remains very strong. For me, this reinforces the point: giving away content for free does not significantly help if you want to gain subscribers. Why would a rational user pay if they can access not only the news/topics mix (homepage) but almost all content for free? And the hope of “locking in” a user through free content apparently does not work broadly. The benchmark clearly shows: only a slight content restriction, combined with effective marketing and attractive trial offers, actually works.

Andreas Demuth is Managing Partner at Upscore.