As others have already pointed out, it's preferable to use element IDs or classes rather than XPath, since XPath can easily break with even minor page modifications (e.g., temporary messages indicating downtime - which happens quite often!).

Additionally, be cautious with browser navigation actions like "back" or "forward." These often lead to unexpected behavior and can disrupt the session on the Handelsregister website, causing XPath selectors to fail because you're no longer on the intended page.

Ultimately, we found it easier to avoid scraping Handelsregister directly and instead switched to structured data providers like handelsregister.ai

Good luck!