Insights Strategy Technology

Third-Party, Direct Or In-House: Which Clean Room Is Right For You?

By St.Clair Mclean on January 4, 2023

Stay up to date

Back to main Blog
St.Clair Mclean

SVP, Infrastructure and Security

Originally featured on AdExchanger 

With more eyes on consumer privacy than ever before, data clean rooms have become one of the hottest technologies in marketing when it comes to secure, modern data collection. 


But as clean rooms come to the forefront, many brands are finding that building and working with this technology requires a little acumen and a lot of planning. After all, not every brand or agency needs or wants the same things. 


Here are three clean room approaches to consider as brands look to align the right capabilities with their needs.


Third-party: easy onboarding and repeatable success

The most familiar type of clean room is the third-party variety. In this model, a brand pays an outside party, like Snowflake, Karlsgate or others, to operate the technology. These clean rooms serve as a middle layer that enables brands, data partners, agencies and platforms to bring in their data. Everyone agrees on the layout, security and match keys. The parties don’t swap PII data, but instead match and share data attributes for enrichment.  


Widespread standardization has made this a quick-to-launch and dependable model, which is why third-party offerings proliferate. The standardization also makes it easy to collaborate with many partners via one platform.


Even though these are off-the-shelf, ready-to-use products, brands must ensure their technology partners match their own security standards. It’s critical to understand what will happen to the data in transit once it leaves the clean room. Is it encrypted? Is it sitting in one place for a period of time? If something goes wrong, who is responsible for the data? These are the basic questions that need to be answered.


Direct relationship: deeper analytics and greater control

The direct clean room represents a more sophisticated model for brands that want greater control over their data, as well as tighter access restrictions. In this model, two partners establish a clean room together. One party spins it up and is the owner of the data, while the other partner can bring its own set of data and/or a set of tools and analytics software to engage with the data. Data sets still need to be matched on an agreed-upon match key to facilitate collaboration. To avoid matching on PII, brands should ensure they have the ability to sync on an anonymous ID.


These kinds of clean rooms are very secure and highly monitored. The clean room and primary data owner should perform security scans on all tools and software brought into the environment. Nothing should flow in or out of the clean room without the partners knowing.


Creating these kinds of environments is generally easy via common data partners such as AWS or Google. But the challenge is to establish the business rules, goals and security standards. It takes more work and negotiation than using an off-the-shelf third-party solution, but the added benefit is tightly controlled security.


Custom built: total control

­­­­­Internal data clean rooms are best suited for brands with designated teams that handle data and analytics. This is especially true for brands that deal with large volumes of data from vendors and agency partners. An internal clean room is the best way to achieve maximum-security standards.


By building internally, brands can create clean rooms that are accessible to anyone in their internal operation. More access to data means that it can be used more widely across an entire marketing and sales operation, which is the ideal of modern data applications.


Of course, a self-contained operation also means brands have the most protection over their data. Nothing comes into the clean room environment, and nothing leaves without the brand’s direct authorization. With these safety standards, the clean room can even be used to protect intellectual property, such as code, analytic methodologies or business intelligence. 


Finding the right fit

Clean rooms enable better data collaboration, execution and measurement with high security and privacy standards. The good news: No brand is limited to just one of these clean room implementations. Test one, employ a combination or use all three. 


Most importantly, maintain high security standards and stay in tune with market changes and technology evolutions. Clean rooms will be an imperative area of expertise for the modern data savvy brand.

Submit a Comment

Stay up to date