It is not necessary. However, I have once studied that the respondents may be varied e.g. pilot can be from one group while the rest from other.

Good Luck,

Regards,

No. However, failing to do so will reduce the relevance of the pilot study to the full research project. This can cause problems if the pilot study is used for things like planning sample size in the main experiment.

Not necessary, you can take separate organisation for pilot study but their (working content, input, process, and product) should be in line with your intended study organisation.

or in other-hand you can take same organisation for pilot study but your pilot study respondents should not be included in the final study.

I hope this suggestion will help you,

Good Luck.