> For clean Markdown of any page, append .md to the page URL.
> For a complete documentation index, see https://docs.pixalate.com/llms.txt.
> For AI client integration (Claude Code, Cursor, etc.), connect to the MCP server at https://docs.pixalate.com/_mcp/server.

# Data feed overlap reports

> Use the Data Feeds overlap reports in Pixalate Analytics to check list download gaps and measure leakage at each probability threshold for Pixalate's pre-bid data feeds.

If you use Pixalate's [pre-bid data feeds](/data-feeds/overview) and Pixalate Analytics, the **Data Feeds** reports show how well those lists work on your traffic. Use them to spot missed downloads and to troubleshoot how a list is applied.

## Open the reports

In the Analytics dashboard, select **Data Feeds** in the left navigation. There's one report for each feed.

| Report                                                                                                  | Feed                                                             |
| ------------------------------------------------------------------------------------------------------- | ---------------------------------------------------------------- |
| IPv4, IPv6                                                                                              | [IP block lists](/data-feeds/ip-block-lists)                     |
| Data Center                                                                                             | [Data center feeds](/data-feeds/data-center)                     |
| Mobile Device ID                                                                                        | [Mobile device IDs](/data-feeds/mobile-device-ids)               |
| CTV Device ID                                                                                           | [CTV device IDs](/data-feeds/ctv-device-ids)                     |
| High Risk Mobile Apps                                                                                   | [High Risk Mobile Apps](/data-feeds/high-risk-mobile-apps)       |
| High Risk CTV Apps                                                                                      | [High Risk CTV Apps](/data-feeds/high-risk-ctv-apps)             |
| High Risk Websites                                                                                      | [High Risk Domains](/data-feeds/high-risk-domains)               |
| Mobile DEFASE Apps, CTV DEFASE Apps                                                                     | [Delisted apps (DEFASE)](/data-feeds/defase)                     |
| Made for Advertising Mobile Apps, CTV Apps and Websites                                                 | [MFA data feeds](/data-feeds/mfa)                                |
| User Agent                                                                                              | [User agents](/data-feeds/user-agents)                           |
| Supply Path Optimization                                                                                | [SPO data feeds](/data-feeds/spo)                                |
| Child Directed Mobile Apps, Child Directed CTV Apps, COPPA Compliance Risk for Mobile Apps and CTV Apps | [COPPA pre-bid data feeds](/compliance/coppa-data-feeds)         |
| Malformed CTV App IDs                                                                                   | [Malformed CTV bundle IDs](/data-feeds/malformed-ctv-bundle-ids) |
| Proxy Gateways                                                                                          | [Proxy gateway feed](/data-feeds/proxy-gateway)                  |

As Pixalate adds data feeds, matching overlap reports are added too.

## Check for missed or late downloads

Each report shows a timeline of when Pixalate published each new list for download and when you downloaded it.

Look for:

* **Missing downloads:** a list Pixalate published that you never downloaded.
* **Long gaps:** a long delay between when a list was published and when you downloaded it.

Either one causes leakage: IPs, device IDs, user agents, apps or domains that are on the current list can get into your traffic because you're blocking with an older list. To fix it, poll for new files on a schedule that matches the feed's update frequency. See [Delivery](/data-feeds/delivery#poll-for-new-files).

## Measure leakage at each probability threshold

Many lists give each entry a probability: Pixalate's assessment of how risky that IP, device ID or other data point is to transact on. Probabilities range from 0.5 to 1, where 1 means the entry is deterministically a risk. You choose the threshold you block at. See [Choose a probability threshold](/data-feeds/ip-block-lists#choose-a-probability-threshold).

The overlap report shows, over a **seven-day window**, how much listed traffic still appeared in your measured impressions at the **0.5**, **0.75** and **1** thresholds. Use it to:

* Check that blocking at your threshold gives the result you expect.
* Decide whether to raise or lower your threshold.

### Lists without probabilities

The Data Center, User Agent and DEFASE lists are deterministic and have no probabilities. For these, the overlap report tells you whether the list is applied correctly: if it is, there should be no leakage at all.

### MFA lists

The MFA overlap is based on traffic from properties rated **medium** or **high** MFA risk. For how the rating works, see [MFA methodology and definitions](/compliance/mfa-methodology).

### CTV app lists

For these lists, overlap is calculated by mapping the bundle or app ID you pass to the app's store ID:

* Made for Advertising CTV Apps
* Child Directed Apps (CTV)
* High COPPA Compliance Risk Apps (CTV)

This affects how you implement these lists for blocking. See [CTV bundle IDs](/mrt/ctv-bundle-ids) for why bundle IDs and store IDs differ.

## Related

* [Data feeds overview](/data-feeds/overview)
* [Data feed delivery](/data-feeds/delivery)
* [Run and customize reports](/analytics/reports)

Questions about a gap or unexpected leakage? Contact [support@pixalate.com](mailto:support@pixalate.com).