An A/B Split test is a marketing experiment that seeks to improve response rates. It involves two kinds of samples: one control sample and one experimental sample. Basically, the experiment sample is the guinea pig to perform different tests on like different subject lines or different copies of text.
The response rates of the experimental sample are going to be compared to that of the control group, in lieu of the changes made and from there a conclusion can be drawn on which sample worked better.
That said, here are a few things to remember with A/B Split testing:
1) One test does not establish a fact – One test result is not conclusive proof. Perform at least four more tests to verify. Once you’ve had an undisputed winner, generate another sample to compare it with so you can draw out which variants worked best.
2) Small changes mean a big difference – One good example for this is testing with a “’free’ downloadable” subject line and a “downloadable ‘free’” subject line. While there are minor differences between these two subject lines in terms of the order of its words, there can be a big difference with regards to their open and response rates, depending on the preference of the respondents/subscribers.
3) Perform one test at a time – To get an idea of what works best for your market, do tests one at a time. You can’t be too sure what works with multiple variables and multiple results all at the same time. Keep yourself focused in one test until you’ve drawn a reasonable conclusion.
4) Give your test enough time – I’ve mentioned this in a previous blog post: Tests take time to be conclusive. Don’t make conclusions after a few hours of performing that test. If you’re testing on behalf of your email marketing campaign, for example, give your subscribers enough time to open or respond to your emails.
5) Focus on Statistical data – Statistical data is an important part of your test given that they give you the measurement you need to judge how well your sample performed. Regardless whether it’s a 1% difference or more, you need to consider how big an impact it left in your campaign and make decisions based off hard numbers that reflect facts.
6) Test regularly – By testing, you think up of new ways to make things work better for both you and your respondents. It’s important to keep testing because with it, you can develop a good habit of finding out the best ways to engage your audience.
For more information on A/B split testing and other facets of Internet marketing, visit Andy Jenkins Blog

Connect at BlogCatalog
Connect at Facebook
Connect at Google+
Connect at LinkedIn
Connect at MySpace
Connect at Pinterest
Connect at Sonico
Connect at Stumbleupon
Follow me at Twitter



























































































