The reliability of the A/B take a look at effects depends on the sample size used. Little samples might cause Untrue positives and negatives, which makes it tricky to conclude In case the dissimilarities are the result of your alterations or random chance. Use rel="canonical": Should you operate a break up take a look at with many URLs, it is best