Varioqub experiments

With Varioqub, you can run experiments on sites of various configurations, ranging from single-page business cards and landing pages to complex sites.

The tool lets you create multi-variant experiments and change not only text, element titles, and colors, but also CSS styles and custom HTML. When the experiment is complete, you can check the metric statistics to see which version of the site users prefer.

To run experiments and split site users into sample groups, Varioqub relies on Yandex Metrica data.

Getting started

Before you start, install a Yandex Metrica tag on your site. Yandex Metrica is required for collecting statistics about sessions and calculating the result.

  1. Prepare your site for an experiment: make changes to your site's code.

  2. Make sure the code works correctly.

  3. Plan out your experiment. Make sure your changes can be applied across all pages and your site is accessible and works properly for users.

    Experiments collect statistics based on actual users' behavior. The more traffic your site receives, the more precise your experiment becomes, helping you spot smaller metric variations. With Yandex Metrica reports, such as Traffic, you can find out how many users visit your site.

  4. Create an experiment.

  5. Review the experiment report and decide on the results.

Data processing

Site user groups are identified

During an experiment, site users are split into groups. Metrics are calculated for these groups to help you evaluate the experiment results.

Site elements are marked up

Use flags to apply changes tested in your experiment. Flags are sets of “key-value” pairs that you can use to change site elements and technical features.

Example of an A/B experiment:

Let's say you want to test if changing a button's color affects conversion rates. The control group sees a white button, while the test group sees a red or blue one. The color choice is passed to your site's code, which then renders the button in a specific color.

How data is processed

  1. A user visits your site.

  2. The site code (frontend or backend) sends a request to Varioqub and passes:

    • Request data (the visitor's IP address, current page URL, and User-Agent).

    • User ID (empty during the user's first session, then retrieved from the _ymab_param cookie for return sessions).

  3. Varioqub returns:

    • User ID to be saved in the _ymab_param cookie.

    • Set of flags that define experimental changes.

    • Technical data for the Yandex Metrica tag.

  4. The technical data is linked to the Yandex Metrica events that the experiment uses for tracking the conversions and displaying report information.

Contact support
Next
Connecting