Comment by BigGreenJorts

This is why sample size is important tho. With a large enough sample size, you can ignore differences in individuals bc the trend of the control will be smaller than the experiment (or not)