Every release can inadvertently break something that was working just yesterday. Automated end-to-end tests check your website just like real visitors do, across multiple browsers and in every environment. This way, you can deploy changes with confidence rather than relying on a hunch.
13.10.2026
Lukas Riesen
When "it's supposed to work" isn't enough
A navigation menu is adjusted, a form gets a new field, a cookie banner is updated. Each of these changes is small. But on a large website, many elements are interconnected: An inconspicuous change can make the search function disappear from the menu or render the contact form unusable on a cell phone.
Manual testing quickly reaches its limits here. No one clicks through every page in three browsers and on two screen sizes before every release. Errors are only noticed when visitors report them.
On top of that, the more frequently you publish updates, the more often you’d need to test. Without automation, testing therefore slows down the very releases that are meant to improve your website—or testing gets skipped altogether due to time constraints.
„With end-to-end testing, you can identify problems before your customers notice them. “
Quick Feedback
The tests run automatically in the CI/CD pipeline and report the results within minutes.
Greater Security
The most important navigation paths through the website are covered in every release.
Less effort
No more manual clicking—your team saves time for more important tasks.
Traceability
Each test run generates a report that even non-technical users can understand.
Here's How We Use End-to-End Tests for Our Clients
Tests that read like requirements
We describe test cases in plain language, for example: "I open the home page and expect the navigation bar to be visible." Behind this is Playwright, a modern tool that controls a real browser and clicks, types, and reads just like a human. The scenarios are written in Cucumber, making them understandable and verifiable even by business users.
Your advantage: You can assess for yourself whether the tests truly cover your requirements.
First, the core functionality; then, the most important processes
We use a two-step process. The quick smoke tests check whether the site is accessible, has a valid certificate, and displays the basic elements. This is followed by the “Happy Paths” tests: the workflows that are most important to your visitors, such as the search function, the FAQ, the contact form, or navigating to your products. Each workflow is tested on both desktop and mobile devices, specifically in Chromium, Firefox, and Safari. This way, we can also catch errors that occur only in a single browser.
Your benefit: Critical functions are protected without the test duration getting out of hand.
The Same Tests in Every Environment
These same tests run locally during development, integration, testing, acceptance, and production. We use tags to selectively choose which tests run and when. After a release to production, we use this approach to verify that everything works live just as it did during acceptance testing. The results are generated automatically: For each browser, you can see how many scenarios passed, were skipped, or failed, along with a link to the detailed HTML report. The notification turns red as soon as something is wrong. If desired, the notification can also be sent to your channel via Microsoft Teams.
Your benefit: Everyone involved can see at a glance whether a release is okay.
End-to-end tests are not an end in themselves!
„These tests ensure that our website continues to do what our visitors expect after every change. Thanks to this new approach, we’re releasing updates more frequently—and with greater confidence!“
Anne Meister
fastforward
Would you like to know how to securely protect your website?
Christian Sprecher Partner, Senior Software Entwickler