Imagine this scenario: you A/B test an idea and it does well. The second test also does well. And the third test ends up outperforming both preceding tests. Now you test the fourth idea and it underperforms. The fifth and sixth ideas also underperform.
Does this mean we need to move on because we’ve discovered the best version of this page? No.
There is nothing known as the best version of a page. You can continue improving a page indefinitely.
When deciding to stop testing ask this question: how important is this page to the overall success of my business? If the answer is very you owe it to yourself to test six more ideas.