How to choose the right number of users for a test?
We advise you on the size of the sample to fit properly the test, ranging from 5 to 50 users. In order to choose the right number of users, you should take into account the following parameters:
- The test effort needs to be proportional to the risk you're taking. The more important the risk is, (ex : modification of a user purchase process of a website with thousands of viewers per month), the bigger the sample size has to be.
- The higher the level of sample targeting is (example: financial manager of big companies), the less the sample size needs to be important to gather reliable results.
- The standard number of testers to gather reliable results during a test with simple criteria level is 30 users. More info on research on research on sample size.
- The higher the number of testers in the sample, the smaller will be the the confidence interval (see graph below).
Choice of the sample according to confidence interval
You can find below recommandations on the number of testeurs according to the level of targeting :
| Simple Examples : age, gender, OS type | Advanced Examples : car brand, job | Customized Example : BtoB targets |
| 30 users | 15 users | 5 users |
