This article applies to Nelio A/B testing versions prior to 5.0.
If you are looking for documentation for our newest version, please bookmark neliosoftware.com/testing/help/
Nelio A/B Testing implements different algorithms for deciding which alternative should see each of your visitors.
- Pure Random. All alternatives have the exact same chance to be seen by any of your visitors.
- Prioritize Original. The original version is more likely to be shown.
- Greedy. Also known as multi-armed bandit, this algorithm makes the best alternative (that is, the one that's converting the best) more likely to appear.
Customers that hold a Professional Plan subscription have the possibility to modify the split testing algorithm used in the experiments of their sites. They just need to open the Advanced tab of the Settings menu item under Nelio A/B Testing menu in WordPress.
Default - Pure Random
That's pure split testing. Each alternative version (including the original one) will get the same amount of traffic. We split your visitors among all the alternatives under test using the same proportion.
Prioritize Original Version
Similar than the random version but specifying a percentage for the original version. Indicate the portion of the visitors that will see the original content. As an example, if you select that 60% of your visitors will see the original content, the other 40% of them will be split among the different alternative versions.
Prioritize Winner (Greedy)
By default, all alternatives in an A/B testing experiment have the same probability of being shown to a user. The Greedy Algorithm changes this behavior and makes sure the vast majority of your visitors see the alternative that is performing better at a time. Exploit the winning alternative and don't loose conversions while testing.