Make a placebo item similar to the thing you are trying to test. For example, if you are testing a new drug such as an anti-migraine pill, make the experimental pills as well as placebo pills that are the same in every regard. The placebo must be the same color, size, shape and have the same markings. Keep the different pills separated so they do not mix at first.
Have a separate researcher, preferably one not connected with the study in any way, place the pills in different containers with markings designating them into group A and group B. The main researcher(s) is not to be told which group is the real pill and which group is the placebo pill. The separate researcher gives the pill containers back to the main researcher(s).
Assign the pills randomly to the experimental subjects. Note which subject is given pills from which group. Perform the experiment as normal and record all the needed observations.
Compile the data into two separate groupings, A and B, and note whether or not the groups proved the experimental hypothesis or disproved it. Only then will the separate researcher reveal which pill group was the placebo and which was the real pill.