How do you deal with A/B testing for small samples?

Question

I am performing A/B testing (basically hypothesis testing) with relatively small samples, so the results are largely inconclusive. I am aware of techniques like CUPED (for decre