Experiments

What is ID?

• ID is an attribute of the HTML element that provides the unique identification of the element among all elements in the web document.

What is Goal?

• Goal is an HTML element (with ID only) that defines the goal action (a click, text editing, etc.); this action is used as indicators of our page conversion (that’s what we track and what data collect); it is also used as statistics for reporting you the best version of your web page.

What is Conversion pixel goal?

• The Conversion pixel goal is a type of goals for A/B/N and multivariate testing that is set in case you have to track the conversion on the websites of some third parties. For instance, to complete the purchase, your customer is redirected to the X resource and has to provide the required information. Thus, only after this step is completed, the purchase is considered to be successful and the conversion is the redirection on the third party website X. So, to track such type of conversion, you may insert the pixel code generated by Maxymyzely and mark your third party website with this pixel.

Can I restore the deleted experience?

• If you have deleted the experiment by mistake, there is a possibility to restore it at once unless you have navigated from the page. Note that only a user with admin rights can delete the experiment.

Why do I need to set goals?

• You set goals to monitor key metrics on your site. (e.g. the Buy button, clicking on the banner, etc.)

What is Content block?

• Content block is an HTML element or a group of HTML elements (ID, XPath, Selector) that has its alternatives (with slight changes) for specific content.

What is Variation?

• Variation is the number of generated alternative variations of the test web page. Variations help to define which pages bring the most conversions of the end-users.

What is Chance to beat all?

• Chance to beat all is the calculation of chances of the page that most probably will win in the current multivariable testing.

What is better A/B/N testing or multivariate testing? 

• A/B/N testing is an individual case of the multivariate testing. If you have more than 2 page versions to compare, use multivariate testing.

What is Conversion rate (CR)?

• It is the percentage of visitors who take the desired action. To be precise, it is the proportion of the taken target actions (desired actions) (A_A) on the website to all generations of our page variations (G_A):

Example: Say, we have an A/B/N experiment with the following components:
Variation A, Variation B and target actions A_A and A_B; so the CR is shown as the proportion of corresponding figures:

What is Error? 

• Error (Sigma) is a deviation from accuracy or correctness in your Conversion Rate calculation.

Why do errors happen? Can I avoid them?

• No, you can’t, but you can mitigate errors by pumping more traffic into your website/app; in plain language it means the more traffic assigned to your experiment – the more visitors do goal actions. More goal actions – more accurate results we receive.

How long should I continue A/B testing?

• You should continue the testing long enough, up to the moment the Chance to Beat All is 95% and higher. Otherwise you won’t get a clear enough picture.