How does exclusive group traffic allocation setup affec experiments allocation setup?
Does the exclusion group traffic allocation overrides every experiment traffic allocation setup included on it?
Or does the exclusion group traffic allocation become the base of every experiment traffic allocation setup? I mean:
Exclusion group config:
Traffic allocation: 50% <<< 50% of 20%?
Traffic allocation: 100% <<< 100% of 30%?
Traffic Allocation and Experiment Eligibility are two entirely different things, decided separately.
If someone is not eligible, they do not count in the traffic allocation decision.
Another way to look at it is that if a visitor meets all of the audience criteria (including any exclusions), they have an x% chance of being included in the experiment. Optimizely does not attempt to track "the last 10 people were all excluded from the experiment, so I need to make the next 5 get into it". The decision is made independently of any other user's decisions.
Or, think of it that every visitor has an x% chance of being included in the experiment. Separately, they must still qualify for the experiment.
So if you have a rule that excludes 30% of your visitors and you have traffic allocation of 50%, you can expect only 50% of the 70% that are not excluded to make it into the experiment (35%).
Analytics and Testing Guru