Article
Newsletters work, but half of your open rate is machines
Two thirds of newsletter subscribers in Germany have already bought or booked straight from one. Yet opens and clicks increasingly come from machines. In five charts, I show which metrics still hold and what you should count instead.
Newsletters work. In a survey of just over 6,000 people in Germany, Austria and Switzerland, 79% had subscribed to at least one newsletter. Two thirds of subscribers in Germany have bought or booked something straight after clicking in a newsletter.
But the numbers most people use to measure their newsletter are becoming less and less reliable. A new analysis by the email provider Omnivery shows that around half of all opens, and now one in six clicks, don’t come from people.
Newsletters generate measurable sales in Germany
The survey was commissioned by United Internet Media, the advertising arm of the German email providers GMX and WEB.DE, in autumn 2025 and published in September 2026. 51% of recipients get newsletters every day, mostly reading them at home and on a smartphone. The most common reason to subscribe is a discount.
For small businesses, coaches and freelancers, this means the newsletter is far from dead. It is one of the few channels where you reach people directly, without an algorithm deciding in between.
The newsletter open rate is half machines
Since autumn 2021, Apple Mail has loaded email content in the background to protect readers from tracking. Each of these preloads counts as an open in your sending tool. According to Omnivery, less than 1% of opens of marketing emails to consumer inboxes were non-human in 2020. Since 2023, it has been around half.
That is why the click long counted as the honest metric. This is changing too: up to 2023, around 2% of clicks came from bots. In the second quarter of 2026, it was 16.6%. That matters, because clicks are rare anyway. In Germany, Austria and Switzerland, the click rate in 2025 was 3.3%, according to the email platform Inxmail.
Where your readers read their email changes your click rate
Most bot clicks come from security checks by mailbox providers. They open links in advance to detect fraud. With Outlook and Hotmail, around 30% of clicks come from machines. Gmail stood at around 6% until early 2026 and then jumped to around 22% at the end of February. With GMX and web.de, which are widely used in Germany, it is only 2 to 4%.
So if your click rate suddenly rises, a new check at Gmail may be the reason, not your subject line.
In B2B newsletters, the security filter often does the clicking
Newsletters to company addresses are hit hardest. Many companies route email through security filters that open and check every link. According to Omnivery, 80 to 95% of clicks there come from machines, and have done since 2020.
On top of that, there are archive sites that subscribe to newsletters automatically. Their links are fetched mainly by Meta’s crawler. Include them, and the bot share for consumer inboxes in the same quarter rises from 16.6 to 58%.
How reliable are these figures? Omnivery sells bot detection itself and has an interest in high numbers. But the report clearly separates consumer and company inboxes and describes its method. The data comes from 126 mailbox providers worldwide, not just Germany. That is why I only use the report’s rounded values.
People recognise your newsletter by its sender
What people actually do before opening is shown, again, by the survey from Germany, Austria and Switzerland. 85% of respondents in Germany check whether the sender looks trustworthy. 79% of subscribers like a consistent, recognisable design, and 73% of respondents recognise the sender faster by its logo.
That is a question of corporate design: the same sender name, the same colours, the same logo, every week. Being recognisable keeps you from being mistaken for spam or fraud in the inbox.
What to count instead
- Count replies and enquiries. A reply by email, a booked appointment or an enquiry through your form comes from a person. Ask a question in your newsletter that people enjoy answering.
- Measure on your website. Add a tag to every link in your newsletter so your analytics show who actually does something afterwards. A scanner opens a link, but it doesn’t book an appointment.
- Don’t remove anyone just for not opening. People reading with Apple Mail seem to open almost always, people who block images seem to never open. Before cleaning up, check whether someone has replied, booked or bought in recent months.
- Separate company and consumer addresses. If a company address clicks every link a few seconds after sending, it was almost certainly a filter. Interpret B2B clicks with care.
- Be instantly recognisable in the inbox. The same sender name, your own domain instead of a free email address, and a fixed design with your logo.
I build newsletters and automated sequences focused on enquiries rather than opens, and corporate designs that people recognise in their inbox. Why newsletters reach more people than many social networks is explained in the article on newsletter reach. If you want to know what your numbers really say, write to me.
Sources
- Omnivery (2026): The State of Email Bots 2026, research report, 10 Aug 2026, B2C campaigns to 126 mailbox providers, 2020 to first half of 2026. omnivery.com
- United Internet Media, Omnicom Media Germany (2026): Erfolgsfaktor Newsletter. Nutzung, Erwartungen und Trends, online survey of 6,014 people aged 18 to 69 in Germany, Austria and Switzerland, 23 Oct to 4 Nov 2025. united-internet-media.de
- GMX and WEB.DE (2026): Rabatte im Postfach. Zwei Drittel bestellen direkt aus Newslettern, press release, 15 Sep 2026. newsroom.gmx.net
- Inxmail (2026): E-Mail-Marketing-Benchmark 2026, Freiburg, 15 Apr 2026, around 4 billion emails to recipients in Germany, Austria and Switzerland in 2025. inxmail.de
