Describes the things that you need to consider while setting up Testing in VWO.
- To run tests in VWO, cookies must be enabled in your browser
- The duration for which you need to run your test depends on several factors such as- number of visitors, conversion rate, etc.
- It is recommended not to change the traffic distribution while a test is running as it can sometimes invalidate your test results regardless of the statistical method that is being used to calculate the results. To know more, refer to Can I Change the Visitor Traffic While a Test is Running in VWO?.
- We recommend not to run a split URL test and A/B test on the same URL at the same time, as it might create issues since the Split URL test will try to redirect the visitor to variation URL while the A/B test would show variation only on control URL. This will cause a discrepancy in the data collected in either of the tests and only result in wastage of visitor quota.
- We recommend not to run multiple split tests or a combination of A/B test and Split URL test on the control URL of your website. Running multiple tests on a control URL causes conflict while executing the test. VWO will send all the test settings during page load so there are chances that visitors’ might become part of only one test depending upon the split URL test settings that are received first.
- Selecting a hypothesis for your test is optional.
- Testing one hypothesis at a time on a funnel is recommended so that there are no interaction effects (where the variation from one test influences the other tests).
NOTE: You can use mutually exclusive features to run multiple experiments
at the same time.