> ## Documentation Index
> Fetch the complete documentation index at: https://docs.flagsync.com/llms.txt
> Use this file to discover all available pages before exploring further.

# Creating Experiments

> Set up and configure experiments in FlagSync to test feature variations.

## Experiment Creation

Set up an experiment in FlagSync to test feature flag variants against metrics.

This process integrates [Feature Flags](/flags/overview), [Events](/events/overview), and [Metrics](/metrics/overview) to evaluate performance and optimize your product with data.

<Steps>
  <Step title="Experiment Details">
    Define a name (e.g., "Button Text Test") and hypothesis (e.g., "Changing 'Buy Now' to 'Complete Purchase' will boost conversions due to clarity").

    <Frame>
      <img alt="Experiment details form" src="https://mintcdn.com/flagsync/dTCxWdoh9aH45uVC/images/experiment-detailed.details.webp?fit=max&auto=format&n=dTCxWdoh9aH45uVC&q=85&s=5d51f6b40e928a27577930d9dac439b1" width="548" height="282" data-path="images/experiment-detailed.details.webp" />
    </Frame>

    <Tip>
      A specific hypothesis aids result interpretation.
    </Tip>
  </Step>

  <Step title="Select Metrics">
    Pick existing [Metrics](/metrics/overview) (e.g., conversion rates, totals) or create new ones, to measure success.

    <Frame>
      <img alt="Experiment metrics form" src="https://mintcdn.com/flagsync/dTCxWdoh9aH45uVC/images/experiment-detailed.metrics.webp?fit=max&auto=format&n=dTCxWdoh9aH45uVC&q=85&s=9f88d1cb6a6902cc11ce95c717ca5d81" width="551" height="205" data-path="images/experiment-detailed.metrics.webp" />
    </Frame>
  </Step>

  <Step title="Choose Feature Flag">
    Select or create a flag with variants to test (e.g., "Buy Now" vs. "Add to Cart").

    <Frame>
      <img alt="Experiment flag form" src="https://mintcdn.com/flagsync/dTCxWdoh9aH45uVC/images/experiment-detailed.choose-flag.webp?fit=max&auto=format&n=dTCxWdoh9aH45uVC&q=85&s=b026f06d1cbb318d1c9c8032cde66e49" width="566" height="504" data-path="images/experiment-detailed.choose-flag.webp" />
    </Frame>
  </Step>

  <Step title="Choose Baseline">
    Pick the control variant (e.g., current "0% discount") to compare against others.

    <Frame>
      <img alt="Experiment baseline form" src="https://mintcdn.com/flagsync/dTCxWdoh9aH45uVC/images/experiment-detailed.baseline.webp?fit=max&auto=format&n=dTCxWdoh9aH45uVC&q=85&s=38755b899045b8ea634f13e3128d3b2d" width="564" height="130" data-path="images/experiment-detailed.baseline.webp" />
    </Frame>

    <Info>
      Typically the current production value, or best-performing variant.
    </Info>
  </Step>

  <Step title="Configure Rollout">
    Allocate percentages to each variant (e.g., 50% baseline, 50% new)—must total 100%.

    <Frame>
      <img alt="Experiment rollout form" src="https://mintcdn.com/flagsync/dTCxWdoh9aH45uVC/images/experiment-detailed.rollout.webp?fit=max&auto=format&n=dTCxWdoh9aH45uVC&q=85&s=ab54af99e370dfd8546abb3e75ca6fa1" width="563" height="308" data-path="images/experiment-detailed.rollout.webp" />
    </Frame>

    <Warning>
      Changing traffic or variants starts a new experiment iteration for data integrity.
    </Warning>
  </Step>

  <Step title="Activation">
    * Save and start the experiment to collect data
    * Once active, FlagSync tracks impressions and events, aggregating metrics to compare variants.

    <Tip>
      Run for a fixed time (e.g., two weeks) or until significant statistical results are achieved.
    </Tip>

    <Warning>
      FlagSync locks linked flags and metrics while running.
    </Warning>
  </Step>
</Steps>

## Next Steps

Get started with a guided flow in [Quickstart: Overview](/quickstarts/overview), then:

* Learn about metrics in [Metrics](/metrics/overview).
* Monitor results in [Monitoring and Analysis](/experiments/monitoring).