What is a data clean room?
Data clean rooms: What exactly are they?
Mostly used for ad targeting, data enrichment, measurement, and audience overlapping, a data clean room is designed to be a secure, neutral, and protected environment where multiple parties can unify and jointly analyze their data. In short, user level data is sent into a data clean room by numerous parties, it gets aggregated in the secure space, and the resulting data is fed back out as a cohort.
This enables businesses and marketers to access high volumes of data while ensuring they remain fully in line with user privacy requirements and that the data they receive is compliant. Perhaps most interestingly, data clean rooms offer a potential way to gain meaningful marketing insights in an increasingly complex measurement and attribution space.
Google was the first to release a data clean room solution with Ads Data Hub in 2017. Shortly after, Facebook followed suit. Next up was Amazon, which launched Marketing Cloud in 2019. By now, there are countless businesses offering data clean rooms.
What is the purpose of using a data clean room?
The goal of a data clean room is to help marketers better understand their data in a joint context while allowing all parties to maintain full ownership over the data they bring to the table. For example, a data clean room could potentially help you identify anything from wasted ad spend or how to take a multi-channel approach to marketing.
They function as a neutral space for first-party data that can be analyzed and leveraged by the multiple parties with access. But how exactly is this kept data privacy compliant? This is an incredibly important question to ask in the GDPR, ATT/iOS 14.5+, Google Privacy Sandbox on Android, and overall increasingly privacy-driven mobile advertising ecosystem. So, let’s explain.
A data clean room is an encrypted, secure location where first-party data is anonymized, layered, and matched with the data from the clean room provider or the inputs from the other parties. This means that no personally identifying information (like a device ID) is available for anyone to view, and this is what makes them an enticing area of exploration for mobile advertisers.
Data that goes in is anonymized and cohorted based on commonalities that are identified from the other inputs fed in but no party ever receives first-party information they didn’t provide in the first place. Data clean rooms are efficient in this manner because they were effectively born out of the privacy-centric movement of the mobile industry, as their ultimate goal is cross-platform compliant measurement.
How does a data clean room work?
There are a few types of data clean rooms, but the common thread is that they are all secure, protected environments that allow multiple data sources to be unified in one space. Data clean rooms typically enable companies to measure performance within their own closed ecosystem.
For example, if we look at Google’s Ads Data Hub, it’s most effective when running campaigns on multiple Google platforms (Search, Display, YouTube), and if you have a substantial amount of first party data to bring to the table, like CRM data.
In summary, data clean rooms work like this:
- Two entities (e.g. an advertiser and a publisher) prepare packages of data and upload them into the data clean room.
- The audiences of both parties are combined and matched in the clean room environment using a one-way hash of identifiers, e.g. email addresses.
- Privacy techniques such as pseudonymization are used to ensure that personally identifying information (PII) associated with an advertiser’s customers is never shared outside of the DCR environment.
- The result is a form of cohorts and aggregated reports—only aggregated data can be displayed, e.g. ‘How many people did X’ as opposed to individual data ‘Who did X’.
How do data clean rooms work in the MMP space and what is Adjust’s perspective?
Channel-agnostic data clean rooms (where an MMP would come into play) can theoretically get a holistic view of performance across channels, but only if publishers feed user level data from their side into the MMP’s data clean room, as this data would be required if we want to combine it with the advertiser’s own user level data and achieve the purpose of the data clean room in the first place.
While Adjust is extremely interested in data clean rooms and their potential to develop further, for now and in the foreseeable future we have determined that they realistically function as a source of aggregated reporting and that they provide data outputs that aren’t functionally able to be leveraged for campaign optimization. And it’s campaign optimization technologies and solutions where we are truly investing for our clients.
To learn more about the Adjust data and analytics suite and to see how we’re empowering clients to stay on top of changes with next generation measurement and analytics, check out Datascape or request a demo today.
Craving monthly app insights? Subscribe to our newsletter.