A/B Testing


Every website or a web application has a goal of getting more users than just visitors. This rate of converting visitors into user is called as “Conversion Rate”. A/B Testing is the process of creating and running two or more variations of a webpage to see which performs best and gets more conversion rate.

When you decide to test your page using A/B Testing, you can test everything on it like header, footer, controls, images, videos etc. You can create multiple versions of a home page by rearranging, removing, adding the UI elements. Running an AB test that directly compares a variation against a current experience lets you ask focused questions about changes to your website or app, and then collect data about the impact of that change.


If you have a website you need to sure that the users like it in terms for user interface functionality and performance. One of our projects needs the analysis of the conversion rate of the user. It was having a registration page with lots of fields, and this lengthy registration process was the huddle for converting visitors to user, so we decided to create different variations for the registration page with minimum fields.

How it Works:

There are many tools that can be used for the A/B testing but we used optimizely (https://app.optimizely.com/) for our use.

Consider a scenario where you want to A/B Test a registration page for its conversion rate. Following steps will explain how to create different variations for the.

1. Create account on the https://app.optimizely.com/ and login, this will redirect you to dashboard page.

2. After this you can see the dashboard where you can find the “New Experiment” button to create new variation for the particular page.

3. Add your experiment name and the target page URL for which you want to create the variations. Please note that URL should not be the local one.

Experiment Name: TestAB.
Experiment URL: www.abregistration.com

4. “Editor” will open the variation editor for creating different variation pages from the URL provided for experience.

This is the original version of the registration page. You can edit the page by right clicking on the page were one can edit the page.

You can remove the element, rearrange it, or classes and html can be edited. One can insert more control from this. Following images show the different variations of the original page

Variation 1:

Variation 2:

5.  Now your variations are ready you can go to settings in the dashboard and get the script url for the experiment.

6. Copy highlighted Script and paste it into <head> tag of your pages which will run under this experiment.

7. Start Experiment.

Routing your traffic to different variations:

Now the important part of this all testing experiment comes where you have to divide the traffic of your web page between these variations. Traffic is randomly assigned to each page variant based upon a predetermined weighting – for example, if you are running a test with 2 page variants, you might split the traffic 33.33/33.33/33.33 or 40/40/20. Visitors are typically cookied so that they will always see the same version of the page (to maintain the integrity of the test). The main factor that decides how much weight you would ascribe to your page variants during a test is timing – whether you are starting the test with multiple variants at the same time or testing new ideas against an established page.

Result and Conversion rates for the Variation:
After successfully running the experience the result can be seen from the dashboard which widely shows the unique visitors, there engagement, conversion rate from different variations..




Michael Patterson sat down with the CEO of Boston Byte, Mustapha Shaikh to discuss the significance and rapid digitization of the healthcar...