Dear miguelm62125791, I'm sorry to say your reasoning is not correct.
The science of AB-testing is complex, nuanced and highly-specific;
without going into too much detail all I can tell you is that even if
the UI shows 99.9% confidence it doesnt mean absolutely anything (as in,
it doesnt mean that you have a 0.01% false-positive rate) if the minimum
sample size has not been achieved.