Frequently Asked Questions

Date of slack thread: 5/27/24

Anonymous: Hey team, I see the product is evolving pretty fast. Do you have plans to allow experiments for web pages to be set up by adding the code once and then managing activation through an admin panel instead of the source code? Editing experiment IDs for new pages in a source code each time can be messy and error-prone. Most A/B testing tools for web pages experiments use a single code snippet and an admin panel to define on which URLs to activate it. This is much easier, especially for landing/web pages experiments.

Anonymous: The issue with layer_ids is that you can’t run overlapping experiments this way. I know that you have plans to add a feature that will account for interactions for overlapping experiments, so being able to use it for web pages experiments would be handy.

Cooper Reid (Statsig): The type of tools you described as for “Web Experimentation”, which solve a specific niche under the experimentation umbrella. These tools are commonly designed for non-technical teams that want to experiment on the top-of-funnel / acquisition side of their business. We offer a platform and solutions that allow Statsig customers to experiment across their entire stack — from web to mobile to backend testing — and are integrating various data sources and their data warehouses. We do have a newer product called Sidecar that addresses that specific use-case you described.