A clean room is a clean room in which work is carried out without external contamination. If you transfer this principle to the digital world, you get a data clean room. What exactly is such a data clean room and why is it important to process some data in it?

Would you like to process sensitive data such as personal data, correlate it with other data or perhaps even link it to third-party data sets? Then you need a solution that guarantees that nobody misuses the sensitive data. For example, personal data cannot be shared with external parties, information that can be used to identify someone cannot simply be linked to an external data set, and certain data cannot even leave the borders of the EU. If you still want to extract value from data in this context, you need a data clean room.

A data clean room is not a room or even a server, but a framework. The reference to a physical cleanroom exists not just because it sounds good in the marketing brochures, but also because the physical cleanroom and the data cleanroom have similar principles. In both cases, the clean room guarantees a safe and strictly controlled environment in which safe activities take place.

In particular for personal data

You usually encounter a data clean room in connection with Personal Data (PII) or simply personal data in Dutch. Especially within Europe, you have to be careful with such data thanks to the GDPR legislation. For example, you cannot simply collect them in a database and then resell them in their entirety to the highest bidder.

Advertisers are of course allowed to offer personalized advertising. This means that they must somehow have access to data based on at least a database of personal data. A data clean room offers a solution here.

A data cleanroom is a secure digital environment that anonymizes personal data and aggregates it according to privacy and compliance rules.

A data cleanroom is a secure digital environment that anonymizes personal data and aggregates it according to privacy and compliance rules. The anonymized data is then available in the clean room under strict conditions for external parties who, thanks to the clean room principle, never have access to the sensitive source data.

Farmer is looking for a wife

In a clean room, a company with access to exciting personal data can, for example, draw anonymous conclusions about the preferences of certain target groups. A digital TV provider knows how fanatical you look Farmer is looking for a wife sees, but must not say so to an agricultural matchmaking website that wants to send targeted advertising.

The provider can anonymize the data and draw conclusions based on additional data. Maybe fifty percent of men between the ages of 31 and 38 watch it regularly Farmer is looking for a wife and perhaps the program is overly popular in the Kempen and Haspengouwse fruit region. A digital advertising platform can use this data to determine when it is statistically interesting to show someone an advertisement for the agricultural dating site (when all indications are that a 35-year-old man from Geel is surfing) and when it is not.

Relevant but not sensitive data

The owner of the personal data encrypts and anonymizes it on the way to the clean room. There the data is processed and divided into important categories. External parties are given access to the clean room where they can test their data against the insights based on the anonymous personal data. At no time does the actual personally identifiable information leave the original collector’s servers or be viewed by any third party.

The clean room contains sufficient relevant data to carry out analyses. Trends become visible and (future) users can be assigned to user categories based on certain characteristics (wealthy car fanatic, young wellness enthusiast, thirsty beer connoisseur…). External parties can combine their data with relevant insights from other parties without anyone throwing away their collected personal data or otherwise violating the GDPR.

Cookies and the data cloud

Cookies are gradually being pushed towards exit and compliance legislation for PII is in place. Against this background, data clean rooms offer the solution to enable collaborations between advertisers on the one hand and content or service providers on the other hand.

Data cleanrooms are essential to a modern data strategy, where everyone retains ownership of collected data but can still share or sell insights to third parties.

One of the parties that focuses heavily on the value of a data cleanroom is, unsurprisingly, Snowflake. After all, this company wants a species data cloud in which companies of all kinds combine their data with that of others (for a fee or free of charge) in order to arrive at new insights. Data protection plays an important role in this.

telecom

At Mobile World Congress, we have a brief chat with Alison Tierney, GVP EMEA at Snowflake, on the sidelines of the Telecom Data Cloud launch. With this offer, Snowflake wants to integrate Telekom players into the data cloud. You have very valuable information for third parties.

For example, you know from your smartphone where you were, when and for how long and when it connects to certain masts. Retailers in a city don’t necessarily want (and don’t need to) know when you’re in the area, but trends about crowds and the presence of certain target groups are of course very interesting.

“Data cleanrooms provide all this population data with no specific data,” Tierney explains again. “Anonymous data is correlated in a clean room, allowing organizations to gain insights based on datasets they don’t own and will never access. This is particularly popular in Europe.”

Clean rooms for everyone

A data cleanroom is so much more than a solution for the post-cookie and GDPR era. A good implementation gives many more companies from many more industries the opportunity to make data available. Anonymous and protected data can still contribute to new insights through clean rooms.

The focus of cleanrooms is on insights from personal data, but possible applications go much further. All companies with datasets they do not want to share can still share data and knowledge about a cleanroom and make it available for correlation with other datasets. It is important to emphasize that everything happens without the sensitive data being copied or ending up somewhere on a remote server outside of the owner’s control. In this respect, a data clean room also fits well into the philosophy of zero-copy integration.

What did we learn today?

Finally, we summarize the essentials: A data clean room is a framework in which you can share insights from sensitive data with third parties without the original data becoming available. The data cleanroom is primarily relevant for companies that want to share insights from personal data with external parties, but the concept is relevant for anyone who wants to remain in control of sensitive data and still share insights.