Building Data-Driven Hypotheses for your A/B Tests

With great power comes, well, a great amount of data to analyze before you make any change to your brand. Yes, of course, responsibility also comes with all of this – but, as data-driven geeks, we just happen to think that the natural order of things is:

  1. Having the power to change something about your eCommerce site;
  2. Analyzing the data behind the site;
  3. Forming a hypothesis based on the data;
  4. Applying changes based on the aforementioned hypothesis.

Doesn’t sound very superhero-ish, we know – but the best superpower humans can achieve is making sound choices, based on real problems and real solutions.

Would Batman ever wear two capes at the same time just because it’s trending?

Nope, probably not. He’d be certain to ask Alfred to run a bunch of statistics on the implications a second cape would have on his credibility with the bad guys. Then, they’d form a hypothesis. And then, they’d test it out.

Or, in better words from our very own superhero, Valentin Radu:

2 Shares
Share2
Tweet
Share