There's no experimentation either without a theory(+), possibly nascent in someone's mind, preferably described systematically in a lab notebook if the research is to go far.
(+) Here, I'm allowing theory to stretch broadly to cover all manner of guesses and hypotheses.