US news publishers lost visitors at a rate of 17.1 percent in the second quarter. German publishers tracked in the same benchmark gained 7.8 percent over the same stretch, a swing of almost 25 percentage points between the two markets. The numbers come from Mather Economics’ June 2026 German News Media Audience Benchmark, published July 21 and covering 51 German brands using Similarweb, Listener, and Sophi.io data. Independent regional publishers sat at the extreme end of that spread, up 30.9 percent year over year.
The easy read of that gap is wrong. Germany is not sitting outside the AI Overviews disruption that has hammered US publisher traffic. Sistrix’s March 2026 data found German position-one click-through falling from 27 percent to 11 percent, a shift the firm links to about 265 million organic clicks disappearing monthly across the market, spread over more than 100 million tracked keywords. That is comparable pressure to what US publishers describe. The divergence in the Mather data is not about whether AI Overviews hit German search traffic. It is about what German audiences did afterward.
Visit frequency, not session depth, is where the two markets split. Germany recorded 2.43 pageviews per visit against 2.47 in the US, a difference small enough to ignore. Frequency told a different story: German readers returned 3.15 times per visitor against 2.14 in the US. They are not reading more per session. They are showing up more often. Large regional group-owned publishers grew pageviews 39.4 percent year over year, and independent regionals added 26.5 percent, both concentrated in the segment where habitual, appointment-style reading runs deepest.
Mather treats audience loyalty and the strength of local newsrooms as likely contributors to that return-visit pattern, stopping short of calling either one a proven driver. The benchmark cites Reuters Institute research showing German readers subscribe digitally at rising rates and report unusually high trust in news brands, offered as supporting context rather than a tested causal link. That restraint matters: national mass-market brands in the same dataset, led by bild.de and t-online, posted the highest frequency in the entire benchmark at 6.58 visits per visitor, which cuts against any simple story where regional-journalism trust alone explains repeat visitation.
The source relationship is worth naming directly. Mather Economics also sells audience and monetisation strategy work to publishers, and a benchmark that lands on engagement durability as the differentiator is also one that supports the case for the kind of consulting Mather offers. That does not make the visitor and pageview counts themselves wrong. It is a reason to weigh the interpretive framing separately from the raw data.
This is a different dataset from the Chartbeat quarterly figures released July 15, which measured how much of publishers’ traffic still arrived via search in each region. Mather’s benchmark is narrower, built around Germany against the US specifically, and organized around engagement rather than referral source. Neither study confirms or undercuts the other; they are measuring different things with different tools.
For a publisher outside Germany, the practical takeaway is a metric, not a market comparison. Track visit frequency per unique visitor as its own line, separate from pageviews per visit. Search referral share shows how exposed a site is to Google’s next AI Overviews expansion. Visit frequency shows how much audience would remain if that channel shrank further. Sites with strong repeat readership, especially regional and niche titles, carry a buffer that referral-share numbers alone will not reveal. Teams that only track session depth and total pageviews should add returning-visitor frequency to the next quarterly review before assuming referral loss is the whole story.
Per a PPC Land report dated July 21, 2026, citing Mather Economics’ June 2026 German News Media Audience Benchmark.
