> For the complete documentation index, see [llms.txt](https://docs.mediarithmics.io/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.mediarithmics.io/collaboration/collaborate-with-a-partner/analyze-cross-partner-behavior.md).

# Analyze cross-partner behavior

Match your dataset with your partner's to compute overlap, and read the resulting insights.

## Your goal

Find out how your audience overlaps with a partner's, and get insights on the common users, without either of you exposing raw data.

## Prerequisites

* An active [Clean Room](/collaboration/plan-and-integrate/setup-clean-rooms/clean-rooms.md) with your partner.
* At least one dataset [shared](/collaboration/prepare-your-environment/share-your-datasets.md) by you and one shared by your partner into that Clean Room.

## Datasets matching

A **datasets matching** is the operation that overlaps two datasets to identify their common records - for example, common identifiers or users.

{% hint style="info" %}
A matching may already exist before you create one. When the [Clean Room Template](/collaboration/prepare-your-environment/manage-your-clean-room-templates.md) behind your Clean Room declares reference datasets, the platform automatically matches every compatible dataset your partner shares against them - see [Auto-matching](/collaboration/prepare-your-environment/manage-your-clean-room-templates.md#auto-matching). Use the steps below to create any other matching yourself.
{% endhint %}

{% stepper %}
{% step %}

### Start a matching

Click **Match datasets** in the Clean Room where datasets have been shared.
{% endstep %}

{% step %}

### Name it

Enter a **Name** that clearly describes the purpose of the matching. A **Description** is optional.
{% endstep %}

{% step %}

### Pick the datasets

Select the datasets to match - one from you and one from your partner.
{% endstep %}

{% step %}

### Choose matching keys and freshness

Select one or multiple [matching keys](/collaboration/prepare-your-environment/share-your-datasets.md#matching-keys) if both datasets are **Segment** type - a matching key is selected automatically when at least one **File** type dataset is involved. Then specify the **identifier freshness** that controls how recent online identifiers must be to be matched (this only applies to identifiers such as cookie, mobile ID, or device ID attached to a Segment-type dataset).
{% endstep %}

{% step %}

### Optionally schedule a refresh

Toggle **Scheduled Refresh** and choose a frequency - **Daily**, **Weekly**, **Fortnightly**, or **Monthly**.

{% hint style="warning" %}
Credits are consumed for any scheduled refresh.
{% endhint %}
{% endstep %}

{% step %}

### Run the matching

Click **Match datasets**. The matching runs immediately; results are typically available within a minute, though processing time varies with dataset size. Track progress in the **Status** column.
{% endstep %}
{% endstepper %}

### Scheduled refresh

When a matching is refreshed - manually or on schedule - the platform recomputes overlap metrics (count and rate), matching stats (total identifiers in the owner's environment and common identifiers), and pushes any new records or deletions to the activation platforms already connected to the matching. Track refresh activity in the **Matching History** tab: timestamp, initiator (a user or the scheduled system), the size of both datasets, and the number of common records.

## Reading the result

Once a matching completes, open it from the **Matched datasets** tab to see:

* **General information** - name, description, creation date, the user who triggered it, and expiration date.
* **High-level metrics** - number of matched records, match rate, and refresh schedule.
* **Overview** - information about the datasets used.
* **Matching history** - the audit trail described above.
* **Matching stats** - per identifier: whether it was used as a matching key, how many common identifiers exist in both datasets, and how many are available for common users in your own environment.
* **Insights** - dashboards on the match, described below.
* **Activation** - feeds available to activate common or exclusive users; see [Activate audiences across partners](/collaboration/collaborate-with-a-partner/activate-audiences-across-partners.md).

{% hint style="success" %}
Every aggregated figure shown here is differentially private - see [Security and privacy](/collaboration/plan-and-integrate/setup-clean-rooms/security-and-privacy.md).
{% endhint %}

## Insights

### Match rate

The first metric to review is the **match rate**. Two views exist, depending on the matching setup:

* **Unitary match rate** - appears when matching one Segment dataset against another Segment dataset. It is the percentage of common users or records between the two, based on the chosen matching keys, together with the absolute matched count. You can switch the comparison to show the rate relative to your partner's dataset instead of yours.

  <div data-gb-custom-block data-tag="hint" data-style="warning" class="hint hint-warning"><p>If at least one dataset is sourced from a datamart, the platform counts matched <strong>users</strong>. If both are sourced from a file or a data warehouse, it counts matched <strong>records</strong>.</p></div>
* **Bulk match rate** - appears when matching a Segment dataset against a Segment catalog. Results are shown per segment in the catalog rather than for the catalog as a whole, including each segment's match rate and absolute matched-user count.

### Dashboards

Dashboards reporting on shared properties of common users between you and your partner can be created by the mediarithmics team on request.

{% hint style="info" %}
The Insights tab is only available when the datasets used in the matching were shared with the **Insights** purpose. Dashboards are currently available in the managed version only - contact your Account manager to request one.
{% endhint %}

## Troubleshooting

**Why can't I see an Insights or Activation tab?** The tab is only available when the datasets used in the matching are Segment type and have been shared with the corresponding purpose (Insights or Activation).

**How long does a matching take?** Typically under a minute, though larger datasets take longer.

## Related pages

<table data-view="cards" data-search="false"><thead><tr><th></th><th></th><th data-hidden data-card-target data-type="content-ref"></th></tr></thead><tbody><tr><td><strong>Activate audiences across partners</strong></td><td>Send common or exclusive matched users to your activation platforms.</td><td><a href="/collaboration/collaborate-with-a-partner/activate-audiences-across-partners.md">Activate audiences across partners</a></td></tr><tr><td><strong>Security and privacy</strong></td><td>Why every figure above is a differentially private approximation.</td><td><a href="/collaboration/plan-and-integrate/setup-clean-rooms/security-and-privacy.md">Security and privacy</a></td></tr></tbody></table>


---

# Agent Instructions
This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com.

## Querying This Documentation
If you need additional information that is not directly available in this page, you can query the documentation by asking a question.

Perform an HTTP GET request on the following URL with the `ask` and `goal` query parameters:

```
GET https://docs.mediarithmics.io/collaboration/collaborate-with-a-partner/analyze-cross-partner-behavior.md?ask=<question>&goal=<user_goal>
```

`ask` is the immediate question: it should be specific, self-contained, and written in natural language.
`goal` is what the user is ultimately trying to achieve, the reason they need the answer. Sharing it helps GitBook give you a better, more relevant answer. A goal is most helpful when it describes the outcome the user wants rather than restating the question. For example, with `ask=how do I create an API token`, a goal like `automate deployments from our CI pipeline` lets GitBook tailor the answer to that use case.

The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.