Skip to main content

​
For Contributors

You can create a partnership with any organisation that is onboarded on Zeotap CDP and has enabled collaboration consent. Only consenting organisations appear in the partner selection dropdown. If you cannot find your intended partner, ask them to contact their Zeotap account team.
Yes. You can create as many partnerships as you need, even with the same subscriber organisation. This is useful if you want to share different attribute sets, or apply different privacy settings, for different use cases.
1,000. This is a fixed lower bound and cannot be reduced. Zeotap enforces this minimum to prevent small cohort sizes from being used to infer individual identities.
Yes. Open the partnership, go to the Configure tab, and click Edit Configuration. You can add more attributes, adjust the aggregate threshold, and change activation settings at any time. Note that your Subscriber may need to re-run analysis to see results based on the updated configuration. Updated settings will not apply to any audience that was saved before the edit.
As the Contributor you set up and govern the partnership, but analysis is initiated and viewed by the Subscriber. You have visibility of the activity log (which shows when analysis runs and when configuration is saved) but you do not see the analysis results panel.
This means collaboration consent has not been enabled for your organisation. Contact your Zeotap account team and ask them to enable it.
Yes. In Step 2 of the configuration wizard, if you enable activation you can select individual channels. Your Subscriber can only activate to the channels you have explicitly permitted, regardless of what channels are connected in their own Zeotap org.

​
For Subscribers

The partnership will appear in your Collaborate → Partnerships list with Pending status. You will also receive a notification from Zeotap. Click the partnership name and then Accept Invitation to proceed. Once the Contributor has submitted their configuration, you can map your join keys.
Leave it unmapped. The analysis will still run using the join keys you have successfully mapped. An unmapped join key does not cause an error — it is simply excluded from the identity matching step. Your overlap result may be smaller, but it will still be valid.
No. The clean room architecture ensures you only ever see aggregated counts. You cannot query, download, or export individual-level records from your partner’s dataset.
Your Contributor partner has set an aggregate threshold — a minimum group size for results to be shown. Any cohort with fewer than that many matched users is suppressed. This is a privacy protection and is not a data quality issue. The note in the results panel will confirm when suppression has occurred.
There are two possible reasons:
  1. The overlap for that cohort is below the Contributor’s aggregate threshold (minimum 1,000). Audiences cannot be saved from groups smaller than this threshold.
  2. Your Contributor partner has not enabled activation for this partnership. Contact your partner if you believe activation should be permitted.
Yes. You can run analysis as many times as you need. Each run queries the current state of both organisations’ data. Running analysis again after you or your partner updates your data will give you fresh results.
Typically a few minutes, depending on data volume. The page refreshes automatically when results are ready.
Suppression shows the users in your dataset who were not matched by your partner. Use it when you want to target people your partner has not yet reached — for exclusion audiences, new customer acquisition campaigns, or understanding the non-overlapping portion of your audience.
Only to channels your Contributor partner has explicitly permitted in their activation settings. If a channel is not available in your activation workflow for a particular partnership, your partner has not enabled it. Reach out to them to discuss.
Once you accept, the status correctly moves from Pending to In Setup. It stays In Setup until both sides have submitted their configuration. If you have submitted your join key mapping but the partnership is still In Setup, it means your Contributor partner has not yet completed their configuration on their end. Check with them.

​
General

No. Collaborate is built on a clean room architecture where raw data from neither side is shared with the other. The Contributor’s data is queried in place within a secure BigQuery environment. The Subscriber never receives individual-level records from the Contributor, and vice versa. Only aggregated counts that meet the minimum threshold are returned as results.
Archiving a partnership removes it from the active partnerships list and prevents any further analysis. Both organisations receive an email notification when the partnership is archived. It does not delete your underlying data — your data source and attributes remain intact in your Zeotap org. Any audiences already saved from previous analyses are also unaffected.
Not currently. Each partnership is a one-to-one relationship between a single Contributor and a single Subscriber. If you need to collaborate with multiple partners, create separate partnerships with each.
Yes. Roles are set per partnership. An organisation can be a Contributor in one partnership and a Subscriber in another simultaneously.
Only the Subscriber sees the analysis results panel. The Contributor sets the parameters but does not view results directly. Both sides can see the activity log for the partnership.
Collaborate supports three core use cases:
  • Overlap — Understand how your audience overlaps with a partner’s. Run an overlap analysis to see the size and composition of your shared audience, broken down by attributes such as gender, age group, or purchase category.
  • Activate — Save matched segments as audiences and push them to ad platforms or CRM systems. The Contributor controls which channels are permitted.
  • Suppress — Identify customers in your dataset who were not reached by your partner. Use the non-overlapping population to build exclusion lists, measure incremental reach, or manage ad frequency across channels.
These three use cases map to the Overlap, Insight Dimension, and Suppression sections in the Analysis tab.
This means no identity matches were found between your dataset and your partner’s using the mapped join keys. Common causes are: the two organisations are using different identifier types or different hashing algorithms (for example, one side has Email MD5 and the other has Email SHA256), the subscriber has not mapped the correct attributes, or the underlying audiences genuinely do not overlap. Check your join key mapping and confirm the identifier type and hash format with your partner.

Quick Start

Contributor Guide

Subscriber Guide

Key Concepts

Last modified on June 30, 2026