Introduction
Users often need to view changes made by experiments without those views affecting the overall experiment reports. This document outlines various methods to achieve this.
Methods to View Changes without Affecting Reports
Beta Snippet Query Parameter
The Beta Snippet Query Parameter, _conv_prevent_tracking, disables tracking for users viewing live experiments if either the cookie or query string parameter is found (e.g., https://www.example.com?convert_preview=live&convert_exp=12345&convert_var=1&_conv_prevent_tracking=true). This allows users to disable tracking when checking a live experiment, which is helpful for debugging purposes. This feature is available only for implementations using the beta snippet.
Blocking IP Addresses
The traditional method involves blocking specific IP addresses from being included in the experiment reports. However, this can be cumbersome and not feasible for dynamic IPs or large groups of users. You may read more about it here.
Force URL and Preview
Using Force URL involves users accessing a specific variation by appending a force URL parameter (e.g., https://www.example.com?convert_action=force&convert_exp=12345&convert_var=1). Although this includes the user in the reports, regular preview by manually adding query parameters each time (e.g., https://www.example.com?convert_preview=live&convert_exp=12345&convert_var=1) is also possible; however, it may not be convenient. You may read more about them here.
User Agent Modification
Another method is User Agent Modification, where users modify the user agent to include the keyword "bot" while using a force variation link, preventing them from being logged in the experiment reports. This can be implemented by adjusting browser settings or using a browser extension to modify the user agent string.
Live Preview on Active Experiment
Live Preview on Active Experiment allows users to utilize the live preview feature on an active experiment (e.g., https://www.example.com?convert_preview=live&convert_exp=12345&convert_var=1), bucketing them into the variation without sending data to the reports and continuing to see the variation until changed by another live preview link.