Unreliable results The Phone Number Database consequences of incorrectly applying scope differ per experiment. In some experiments the difference is small, while in other experiments it makes the difference between Phone Number Database a significant improvement or not. Determining the impact is only possible when you also have the correct data. This means that with only an incorrect scope you can never be sure whether Phone Number Database or not there is an effect. This uncertainty means that your results are not reliable. The solution The solution is to collect the data correctly and use a hit scope.
With every interaction you will Phone Number Database have to indicate which experiment is currently active. If you do that, you can choose the scope 'hit' and make sure that you do not link data that you should not link. Scope A B testing with interaction and measurement. Often the standard integrations do not Phone Number Database offer this possibility. You then have to make this integration yourself. You can do this by using the Javascript APIs or cookies from the A/B test packages. You can also keep track of which Phone Number Database experiment is currently active for the user in a cookie. Before sending data, check the active experiments and enrich the interaction data with the information about the active experiments.
In exchange for a more Phone Number Database complex setup, you get correct data, which you can store in dimensions with the scope 'hit'. One last choice In the example above, we send the relevant information with each event. This ensures that we can choose the scope 'hit'. In our reports we will see that all interactions from hit Phone Number Database are linked to variant 1B and the interactions from hit are linked to variant 1B and 2C. However, this leads to a final choice. Do you use one dimension for all experiments Phone Number Database or do you use a separate dimension for each experiment.