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

# Experiments

> Experiments help you evaluate how different marketing strategies perform by analyzing outcomes across randomized experiment groups. You can measure lift, compare treatment and holdout results, and understand whether your campaigns drive meaningful changes in user behavior.

> [!NOTE]
>
> **Experiments** were previously called **Splits**.

> [!NOTE]
>
> Experiments is only available on [Business tier plans](https://hightouch.com/pricing).

## Before you begin

* Created [Audience Experiments](https://docs.hightouch.com/customer-studio/experiments.md) in Customer Studio
* Audience synced at least once after experiments were created
* Event data (such as purchases or clicks) available for measurement

## Overview

**Experiments** provide a measurement layer for [Audience Experiments](https://docs.hightouch.com/customer-studio/experiments.md).

Whenever you create an Audience Experiment in Customer Studio or configure an experiment on an [A/B split tile in Journeys](https://docs.hightouch.com/customer-studio/journeys/tiles.md#ab-split), Hightouch automatically generates a corresponding **Experiment** so you can track performance across randomized groups.

Experiments allow you to:

* Compare holdout vs. treatment outcomes
* Analyze lift and confidence intervals
* Visualize performance over time
* Evaluate strategies continuously

| Feature | Description |
| - | - |
| **Experiment results** | Displays lift, performance trends, and confidence intervals |
| **Configuration** | Controls metrics, measurement windows, and start dates |
| **Normalization** | Enables per-member or baseline-scaled comparisons |

> [!NOTE]
>
> Experiment measurement charts have been updated. All experiment reporting now
> lives in **`Intelligence` → `Experiments`**.

## How Experiments are created

Experiments are automatically managed based on Audience Experiments in Customer Studio:

* Creating an **Audience Experiment** automatically creates a corresponding **Experiment**.
* Disabling or deleting an Audience Experiment removes its Experiment.
* Restoring an Audience Experiment restores its Experiment.

This ensures measurement stays aligned with the audiences you're actively using.

## Setup and requirements

### 1. Create Audience Experiments

Before measuring experiment results:

1. Create [Audience Experiments](https://docs.hightouch.com/customer-studio/experiments.md) in Customer Studio.
2. Ensure the audience has synced at least once *after* the experiment groups were created.
3. Confirm that users generate measurable events (such as purchases, page views, or clicks).

> [!TIP]
>
> For every Audience Experiment created in Customer Studio, Hightouch
> automatically creates a corresponding Experiment in **`Intelligence` →
> `Experiments`**. Deleting or restoring an Audience Experiment removes or
> restores the linked Experiment.

## Measure results

The **`Experiments`** section of Intelligence helps you compare outcomes between experiment groups and evaluate the impact of your campaigns.

Navigate to: **`Intelligence` → [`Experiments`](https://app.hightouch.com/experiments?from=docs)**

![Intelligence navigation menu](https://mintcdn.com/hightouch-3c037897/L_vQfA0JQEpeGUUo/images/campaign-intel/experiments/ci-experiments-nav.png?fit=max\&auto=format\&n=L_vQfA0JQEpeGUUo\&q=85\&s=a7feee8986eeb459347074806413007c)

### 1. View list of Experiments

The Experiments list shows all Experiments, their statuses, and recent updates.

Statuses include:

* **Draft:** Missing one or both required configuration elements (primary metric or start date).
* **Scheduled:** Fully configured; the start date is in the future.
* **Running:** Fully configured; the start date is today or in the past.

  ![Experiments list](https://mintcdn.com/hightouch-3c037897/L_vQfA0JQEpeGUUo/images/campaign-intel/experiments/ci-experiments-overview.png?fit=max\&auto=format\&n=L_vQfA0JQEpeGUUo\&q=85\&s=ecbf831248db8839a514708bdaacc57b)

### 2. Configure an experiment

Open an experiment and select the **`Configuration`** tab.

From here, you can:

* **Choose a primary metric** (required) and optional secondary metrics (e.g., *Conversions*, *Revenue*).
  * [Learn how to create a metric](https://docs.hightouch.com/campaign-intelligence/metrics.md).
  * *Optional:* Use the **`Filter by`** dropdown to refine by user properties or events.
* **Set a Start date**
  * Determines when measurement begins.
  * Does **not** apply retroactively.
  * Does *not* affect sync or activation behavior.
* **Choose a Measurement window**
  * Example: *Entry → 30 days after entry* measures events from the moment the user enters the audience to 30 days later.

    ![Experiment configuration](https://mintcdn.com/hightouch-3c037897/L_vQfA0JQEpeGUUo/images/campaign-intel/experiments/ci-experiments-configure.png?fit=max\&auto=format\&n=L_vQfA0JQEpeGUUo\&q=85\&s=8522cbc6855a06f3dfd98c1a44907ab3)

### 3. Interpret results

The **`Overview`** tab displays experiment outcomes and performance trends.

![Experiment results chart](https://mintcdn.com/hightouch-3c037897/L_vQfA0JQEpeGUUo/images/campaign-intel/experiments/ci-experiments-interpret.png?fit=max\&auto=format\&n=L_vQfA0JQEpeGUUo\&q=85\&s=7b89e7d4e375b365c17f7ba787341431)

Key elements:

* **Lift %:** Percentage difference between treatment and holdout group performance.
* **Lift interval bar:**
  * **Green:** Significant positive lift
  * **Red:** Significant negative lift
  * **Gray:** Not statistically significant (interval overlaps 0%)
* **Performance lines:**
  * Solid lines show performance over time
  * Shaded regions represent 95% confidence intervals

> [!TIP]
>
> Lift intervals use a Bayesian method, enabling continuous monitoring without
> needing to wait for an experiment to complete.

### 4. Normalize results

Use the **`Normalization`** dropdown to switch perspectives:

* **Per member (default):** Average performance per user
* **Normalized to baseline group:** Scales performance to compare groups evenly

Hover over the lift card to view raw totals.

![Normalization toggle](https://mintcdn.com/hightouch-3c037897/L_vQfA0JQEpeGUUo/images/campaign-intel/experiments/ci-experiments-normalize.png?fit=max\&auto=format\&n=L_vQfA0JQEpeGUUo\&q=85\&s=1a77c3eabdcebb601941ae4b6cd1014d)