Why Split Tests Are Like High School

I’m talking about split testing . What is a split test for sex? In fact, many… Let me explain. Everyone says they’re doing it… Just like sex in high school, subtests are all the rage. Everyone likes to pretend they are experts. Buzzwords and rumors abound… stories about changing the color of the checkout button to increase conversions by an order of magnitude (but no one shares the magic color!). Most importantly, no one wants to admit that they really don’t know what they’re doing, or (wow!) have never done it themselves. Many join the conversation and don’t want to get split on, they don’t even know what the test is! Let’s start

Let’s Explain Why With a Simple Example

want it to be at least 90%, preferably 95% above picking the winner. You don’t have to worry about calculating the numbers yourself; there are free tools out there that can calculate the statistical significance of the results for you (you just plug in the number of impressions and actions for each Singapore Phone Number List version, and the rest is done for you), and scores Testing tools like Google Website Optimizer will do the calculations for you, too (and plug in the premise ). If you don’t want to calculate the actual meaning of the test, here’s a thumbs-up you can use (rules for landing page optimization , borrowed from Tim Ashe’s book ): If there are 100 show samples, you need to see a gap between the

Most Don’t Do It Very Well

Singapore Phone Number List
Singapore Phone Number List

In order to effectively split tests, you need to isolate variables, which means when testing one thing! (Okay, yes, it’s technically possible to test multiple things at once – it’s called multivariate testing but in practice, doing it requires huge traffic numbers, a more complicated setup – if you’re not already is doing, then it may not be for you.) Did you measure the results? I mean actual measurements, vs numbers? It’s also a rarity – more often it’s a “measure” like “I think we’re getting more signups”. Be careful with this, because as characters like everyone suffer from confirmation bias , which means we’re more likely to favor evidence that supports what

Leave a comment

Your email address will not be published.