Sample size is too small. Correlation is not causality. Try harder next time.
You're assuming that what they got wasn't their intended result. If it was, they got just want they wanted. No need to try again, eh?