3 ms·
Anytime I see some crazy wording like, "Which of these call-to-action button versions increased clickthroughs by 600%?", my first thought is "Too small of a sam
by ScottWhigham 13y ago
Anytime I see some crazy wording like, "Which of these call-to-action button versions increased clickthroughs by 600%?", my first thought is "Too small of a sample size". Over a large sample size, it's hard to believe that the difference in those two buttons would be anywhere close to 600%. Maybe I missed it - did she say anywhere how many days/impressions were used?
- siddharthdeswal 13y agoSample size was 4500+ distributed between the two versions. In this case, the sample size was not small but the conversion rate of the page was very low. Which is why a 600% increase was even possible. That's usually a problem with A/B testing in the real world. There are so many websites whose conversion rates hover in the 0.xx% range. Thoughtful hypotheses almost always result in significant gains. Not because they did something magical, but because they simply communicated their business offer better.