Website traffic allocation lets you choose the percentage of visitors being included in that experiment. As you allocate the website traffic for an experiment, only that selected portion of visiting will be shown in the experiment.
Consider, you run an AB Testing on a certain page. And you have set the traffic allocation to be 25%. If the total visits to your webpage after the experiment has started is 100, only 25 visitors of the total 100 would be subjected to the experiment, while the rest 75 would see your webpage as it is hosted earlier.
Also, note that traffic allocation is merely allocating the probability that any particular visitor visiting the experiment is subjected to an experiment.
Why need web traffic allocation?
Web Traffic Allocation might come in handy when, You are unsure of the experiment being run and you do not want the user to feel the mishaps. In this case, you can start experimenting with a small number of users (say 10%) and check the results, tweak the experiment and repeat to find the right fit for your needs.
You want to define the percentage of visitors to be shown with a particular variation. In this case, you can specify the traffic percentage of each variation.
Please note that the variations percentage is subjected to the total traffic percentage allocated for that experiment. If you have an experiment with total traffic percentage allocated to be 50% and two variations in the experiment with 50% traffic each and if the total visitors is 100, only 50 visitors will be included in the experiment in which 25 would have seen the first variation while the other 25 would have seen the other variation.
You know that a particular variation in an experiment is winning and want all your visitors henceforth to see that particular winning variation. In this case, you can edit the experiment and set the traffic for that variation to be 100%.
we recommend not to adjust the traffic allocation after the experiment is started. Read how to optimize website traffic