There are tools available to see how long your test should run to give you optimum results. Tools like Duration Calculator and A/B test Calculator take into account parameters such as average visitors or participants, conversion rate and number of variation in the test to help you determine if you need more time (in days) to run your A/B testing or if more variations are required in your testing. Also, they help you understand and find the number of visitors needed to gain the statistics confidence of more than 85%.
Other than that, it is important to follow a few steps to run your tests appropriately:
1. Run your test for your complete business cycle
2. Always consider the variance in your traffic based on the type of days i.e. weekdays, holidays, and weekends.
3. To make your test results statistically valid, don’t stop them until you have the reached the minimum sample size.