More IDs, more reach, more transparency: Introducing Datalab v2

Decentriq's Datalab is the publisher-side data environment that powers secure collaborations in our media and advertising clean rooms. This environment serves as the configuration layer that determines which audiences, identifiers, and collaboration types a publisher makes available for collaborations.

The identity landscape has gotten more complex

Publishers today manage a patchwork of identifiers (e.g. first-party user IDs, hashed emails, MAIDs, and alternative IDs like Utiq, First-id, NetID, or ID5) often all at once and for the same user. Advertisers, meanwhile, increasingly want to bring their own preferred ID type and expect a match to be possible with the one they have chosen.

Until now, Decentriq's Datalab required publishers to pick a single matching ID type when setting up their data. That worked well for simpler setups, but created friction for publishers with richer, more complex identity graphs and left brands guessing about which collaboration types were actually possible before a clean room was even created.

What's new: multiple ID types, one publisher data environment

With Datalab v2, publishers can now configure any combination of identifier types into a single Datalab. Rather than specifying one matching ID at setup, publishers define an Identifiers table that can hold any number of named IDs. These IDs are individually tagged based on their role, whether they are used for matching with an advertiser's seed audience, for activating the resulting audience, or for both.

This means a publisher can bring, for example, a hashed email for matching and a MAID for activation into the same Datalab without needing to create separate configurations or run separate pipelines.

Supported ID types now include:

• Hashed email and phone
• Mobile advertising IDs (MAIDs and hashed MAIDs)
• Alternative IDs: Utiq Martech ID, Utiq Adtech ID, First-id, ID5, NetID
• Custom first-party string identifiers

Publishers also declare upfront which collaboration types a given Datalab is designed to support: Overlap, Remarketing, AI Lookalike audiences, Rule-based audiences, or Insights. This shapes what data is required and what statistics are generated, and makes it immediately clear to brands what kinds of activation are possible.

More transparency for the buy-side

One of the most meaningful changes in Datalab v2 is what it unlocks for brands and agencies on the buy-side.

When a publisher provisions a Datalab v2 into a clean room, brands can now see at a glance not just that a collaboration is possible, but exactly what's available: which identifier types the publisher supports for activation, how many reachable IDs exist for each, and what collaboration types are on offer. This replaces guesswork with concrete, verifiable data all within the same confidential computing environment that keeps the underlying personal data fully private.

For campaigns where reach matters, publishers can also see the activation potential broken down by collaboration type: how many activation IDs can be reached via Remarketing, how many users with segments are addressable for AI Lookalike targeting, and more. The Datalab generates all of this automatically as part of its validation and statistics report with no additional work required from the publisher.

Publishers also retain full control over which activation ID is used for a given collaboration, and can switch their selection at any time. Additionally, brands can see exactly which activation ID has been selected, so there's no ambiguity about what will actually be delivered.

Backwards compatible, and built for what's next

Existing Datalabs continue to work as before, and publishers can take their time migrating to v2 at a pace that suits them. During the transition, both versions are visible in the Decentriq portal, clearly labelled.

For publishers ready to move to v2, the setup wizard walks through the process step by step. Simplified schema formats are available for the most common setups, so publishers with straightforward configurations don't need to change how they format their data.

Get started

Datalab v2 is available now. If you're a publisher looking to expand which ID types you support or make your inventory more transparent to brand partners, reach out to your Decentriq customer success manager or get in touch with our team.