Skip to main content

Documentation Index

Fetch the complete documentation index at: https://docs.ghostlyx.com/llms.txt

Use this file to discover all available pages before exploring further.

​
How GhostlyX fits into A/B testing

GhostlyX does not serve variants or split traffic. Use your own A/B testing tool or feature flag system (Optimizely, LaunchDarkly, GrowthBook, or a custom implementation) to assign visitors to variants. GhostlyX then measures the conversion outcome for each variant.

​
Instrumenting variants

When a visitor is assigned to a variant, fire a custom event with a property that identifies which variant they saw: 
ghostlyx('event', 'Experiment', { variant: 'A' })
For variant B: 
ghostlyx('event', 'Experiment', { variant: 'B' })
You can use any event name and any property name that makes sense for your experiment.

​
Tracking conversions per variant

Fire a second custom event when the conversion you are measuring occurs. For example, if you are testing a pricing page and measuring sign-ups: 
// When the visitor signs up
ghostlyx('event', 'Sign up', { variant: 'A' })

​
Comparing results in the dashboard

  1. Create a Custom event goal for each event name you are using.
  2. View the goal in the dashboard to see completion counts and conversion rates.
  3. Filter the dashboard by variant property value to isolate each group.
Because GhostlyX is a measurement layer, you retain full control over the experiment design and statistical analysis. Use your A/B testing platform’s built-in significance calculator, or export the raw counts and run your own statistical test.

​
Compatible tools

GhostlyX works alongside any client-side or server-side A/B testing tool, including:
  • Optimizely
  • LaunchDarkly
  • GrowthBook
  • VWO
  • Custom feature flag systems