Title: Cohort Management
Locale: en
URL: https://sensorswave.com/en/docs/user-operation/cohort-management/
Description: Learn how to create, manage, and apply cohorts across various scenarios
This guide covers the complete lifecycle management of cohorts, including how to create, edit, use, and delete cohorts, as well as how to apply cohorts in analytics models, Feature Gates, and Experiments.

## Creating Cohorts

Sensors Wave provides four ways to create cohorts, each suited for different scenarios.

### Method 1: Direct Creation

Create a cohort directly from the Cohorts module:

1. Click **Users** > **Cohorts** in the left navigation bar
2. Click the **New Cohort** button in the top right corner
3. Select the cohort type:
   - **Dynamic Cohort**: Saves the rules; users are automatically updated
   - **Static Cohort**: Saves a user snapshot; fixed and unchanged
4. Configure cohort rules (see [Cohort Rules](cohort-rules.mdx))
5. Click **Estimate Size** to confirm the user scale
6. Fill in cohort information:
   - **Cohort Name**: A concise description of the Cohort Definition
   - **Cohort description**: A detailed explanation of the cohort's purpose and rules
   - **Folder**: Select the folder for the cohort
7. Click **Save** to complete creation

### Method 2: Save from User List

Save a cohort after filtering in the User List:

1. Configure filter conditions in the User List
2. Confirm the filter results meet expectations
3. Click the **Save as Cohort** button in the top right corner
4. Select the cohort type:
   - **Dynamic Cohort**: Saves filter conditions as rules
   - **Static Cohort**: Saves the current user list as a snapshot
5. Fill in the Cohort Name and description
6. Click **Confirm** to complete creation

### Method 3: Save from Analytics Models

Save users as a cohort from analytics model results:

**Save from Segmentation**:

1. In Segmentation, use "Unique Users" as the aggregation method
2. In the data table, click a value to enter the user list
3. Click the **Save as Cohort** button
4. Fill in cohort information and complete creation

**Save from Funnel Analysis**:

1. In Funnel Analysis, view the converted or dropped-off users at a step
2. Click the **Save as Cohort** button
3. Choose to save "Converted users" or "Dropped-off users"
4. Fill in cohort information and complete creation

### Method 4: Upload User List

Create a Static Cohort by uploading a file:

1. Click **New Cohort** > **Upload Users**
2. Download the template file and fill in the user ID list
3. Upload the completed file (supports CSV, Excel)
4. The system validates and matches users
5. Confirm the match results, fill in cohort information
6. Click **Save** to complete creation

**Upload file requirements**:

- File format: CSV or Excel (.xlsx)
- Required column: User ID (must match user identifiers in the system)
- Maximum per upload: 100,000 records
- Character encoding: UTF-8

## Cohort List Management

### List Display

The cohort list page displays all created cohorts:

| Column | Description |
|--------|-------------|
| **Cohort Name** | The name of the cohort; click to view details |
| **Cohort type** | Static Cohort or Dynamic Cohort |
| **Size** | The current number of users in the cohort |
| **Created by** | The creator of the cohort |
| **Created at** | The creation time of the cohort |
| **Last updated** | The most recent modification time of the cohort rules |

### Search and Filter

Quickly find the target cohort:

- **Search**: Enter Cohort Name keywords to search
- **Filter by type**: Filter by Static Cohort or Dynamic Cohort
- **Filter by folder**: Filter cohorts by folder
- **Filter by creator**: Filter cohorts by creator

## Editing Cohorts

### Edit Rules

Modify the rules of a Dynamic Cohort:

1. In the cohort list, click the target cohort
2. Click the **Edit** button
3. Modify the cohort rules
4. Click **Estimate Size** to confirm changes
5. Click **Save** to apply modifications

> **Note**: The user list of a Static Cohort cannot be edited. If modifications are needed, please create a new cohort.

### Edit Information

Modify the name, description, and other information of a cohort:

1. On the cohort details page, click the edit icon next to the name or description
2. Modify the content
3. Click **Confirm** to save

### Move Cohorts

Move a cohort to another folder:

1. In the cohort list, select the target cohort
2. Click the **Move to** button
3. Select the target folder
4. Click **Confirm** to complete the move

## Viewing Cohort Details

### Cohort Overview

The cohort details page displays the following information:

- **Basic information**: Name, type, description, creator, creation time
- **Size trend**: A trend chart of the cohort size over time (Dynamic Cohorts)
- **Cohort rules**: The complete rule definition of the cohort
- **Usage**: List of analyses, Feature Gates, and Experiments that reference the cohort

### View User List

View the specific users in a cohort:

1. On the cohort details page, click the **View Users** button
2. Redirects to User List, automatically filtered to the cohort's users
3. You can further filter, view details, or export

### Size Trend

Dynamic Cohorts support viewing historical changes in size:

1. On the cohort details page, view the "Size Trend" chart
2. Select the time range: Past 7 days, 30 days, or 90 days
3. Observe the trend of user count changes

## Deleting Cohorts

### Pre-deletion Check

Before deleting a cohort, the system checks the cohort's references:

- Whether it is referenced by other cohorts
- Whether it is used in analytics models
- Whether it is used in Feature Gates
- Whether it is used in Experiments

If references exist, they must be removed before the cohort can be deleted.

### Execute Deletion

1. In the cohort list, click **More** > **Delete** on the cohort row
2. Confirm the deletion
3. The cohort is deleted and cannot be recovered

> **Note**: Deleting a cohort is irreversible. Please proceed with caution. We recommend moving the cohort to a "To be deleted" folder first and observing for a period before deleting.

## Using Cohorts in Analytics Models

### As a Segment By Condition

Filter specific cohort users in analytics models:

1. Create or edit an analysis (Segmentation, Funnel Analysis, Retention Analysis, etc.)
2. In the **Segment By** area, click **Add Filter**
3. Select **Cohort**
4. Select the operation: **Belongs to** or **Does not belong to**
5. Select the target cohort
6. Execute the query to view filtered results

**Example**: Analyze high-value user behavior

```
Segment By: Belongs to "High-Value Users" cohort
Event: Browse Product
Aggregation: Total count
Group By: Product category
```

### Compare Multiple Cohorts

Compare the performance of different cohorts in Segmentation:

1. Add multiple Segment By filters
2. Segment 1: Belongs to "High-Value Users" cohort
3. Segment 2: Does not belong to "High-Value Users" cohort (regular users)
4. Execute the query and compare the two user groups in the chart

### Use in Group By Dimensions

Group data by cohort:

1. In Group By dimensions, select **Cohort**
2. Select multiple cohorts to compare
3. Execute the query to display data distribution by cohort

## Using Cohorts in Feature Gates

### Configure Targeting Rules

Enable features for specific cohorts:

1. Navigate to the **Feature Gates** module
2. Create or edit a Feature Gate
3. In **Targeting Rules**, click **Add Rule**
4. Select rule type: **Cohort**
5. Select the target cohort
6. Set the gate status: Enabled
7. Save the configuration

**Example**: Enable exclusive features for VIP users

```
Feature Gate: vip_exclusive_feature
Rule 1: Cohort belongs to "VIP Users" → Enabled
Default: Disabled
```

### Combined Usage

Cohorts can be combined with other rules:

```
Feature Gate: new_recommendation_v2
Rule 1: Cohort belongs to "Internal Employees" → Enabled (internal testing)
Rule 2: Cohort belongs to "Beta Users" AND Device = iOS → Enabled (iOS Beta)
Rule 3: Percentage 10% → Enabled (10% rollout)
Default: Disabled
```

See [Feature Gates - Create and Configure](../feature-gates/create-and-configure.mdx) for details.

## Using Cohorts in Experiments

### Configure Allocation Conditions

Run experiments only for specific cohorts:

1. Navigate to the **Experiments** module
2. Create or edit an experiment
3. In **Allocation Strategy**, configure target users
4. Select **Cohort**
5. Select the target cohort
6. Configure experiment groups and ratios

**Example**: Run a new feature experiment only for active users

```
Experiment: new_checkout_flow
Target users: Belongs to "Active Users" cohort
Groups:
  - Control (50%): Original checkout flow
  - Treatment (50%): New checkout flow
```

### Exclude Specific Cohorts

Exclude certain users from experiments:

```
Experiment: pricing_experiment
Target users: All Users
Exclude: Belongs to "Internal Employees" cohort
Exclude: Belongs to "Already in Other Experiments" cohort
```

See [Experiments - Targeting and Allocation](../experiments/targeting-and-allocation.mdx) for details.

## Exporting Cohort Users

### Export Operation

Export the cohort user list as a file:

1. On the cohort details page, click the **Export Users** button
2. Select the export format: CSV or Excel
3. Select the export fields: User ID, User Properties, etc.
4. Click **Confirm** to start exporting
5. Wait for the export to complete and download the file

### Export Limits

- Maximum per export: 100,000 records
- When exceeding the limit, please use batch export
- Sensitive fields may be masked based on permissions

### Use Cases

- Import into CRM systems for user outreach
- Import into advertising platforms to create targeted audiences
- Match with external data for analysis

## Cohort Usage Statistics

### View Usage

Understand how a cohort is being used:

1. On the cohort details page, view the "Usage" area
2. The following reference information is displayed:
   - Which analytics models use it
   - Which Feature Gates use it
   - Which Experiments use it
   - Which other cohorts reference it

### Clean Up Unused Cohorts

Regularly clean up cohorts that are no longer in use:

1. Filter cohorts with empty "Usage"
2. Check whether the cohort still has value
3. Delete or archive confirmed unused cohorts

## Notes

### Permissions

| Operation | Required Permission |
|-----------|-------------------|
| View cohort list | Users - Consumer |
| Create cohorts | Users - Analyst |
| Edit/Delete cohorts | Cohort creator or Manager |
| Export users | Data export permission |

### Dynamic Cohort Performance

- Dynamic Cohorts are computed in real-time each time they are used
- Complex rules may impact query performance
- We recommend periodically optimizing cohort rules

### Cohort Consistency

- Users in Dynamic Cohorts change over time
- Using the same cohort at different times may yield different users
- When you need a fixed set of users, use a Static Cohort

## Next Steps

Now that you understand how to manage and use cohorts, you can:

1. **[Integration with Analytics](integration-with-analytics.mdx)**: Learn more about using cohorts with analytics
2. **[Best Practices](best-practices.mdx)**: Reference cohort design and management guidelines
3. **[FAQ](faq.mdx)**: View answers to frequently asked questions about cohorts

---

**Last updated**: January 19, 2026