You're unlikely to learn anything, statistically conclusive, from 10k views -- unless the test is like 100% better than the control, which is very rare in testing.