The role of data clean rooms in a privacy-first world
- Posted by Davide Rosamilia
- On Aug 03, 2022
Data clean rooms have earned themselves a spot in the ‘hot topics’ category in recent months, but don’t be fooled, they are no new phenomenon to the industry. They have been around for many years but are now taking center stage for one overarching reason: the industry is becoming more and more privacy-focused.
As we move into a world with tightening regulations, with traditional signals being kicked to the curb, and consumers being more interested in what happens to their data, the industry is demanding solutions that allow the continual sharing of data in a safe and privacy-first way. As one of the solutions that the market is adopting in order to operate in cookieless environments, there is an opportunity for data clean rooms to shine.
What are data clean rooms?
Clean rooms are safe data storage spaces that allow for the cross-matching of anonymized, aggregated data by two or more parties for the purpose of digital advertising. They enable secure entry to distributed datasets without actually revealing any personally identifiable information or allowing access to the data outside of the clean room.
The involved parties have complete control over how their data is used, user profiles are kept entirely anonymized, and no first-party data leaves the clean room.
How do data clean rooms work?
Clean rooms involve a series of steps to ensure that data is handled safely and in a privacy-ensured way. These steps include:
- Each party puts their first-party data into the clean room. A number of privacy and security measures are then applied to the data.
- The data can now be activated.
Data clean room uses cases
Data Clean Rooms can be used for:
- Performance measurement – Let’s take an ecommerce brand as an example. A footwear brand is selling a new line of trainers and purchases a certain number of impressions from a news website to reach its target audience. The brand wants to understand whether those ad placements generated positive campaign results. Within the data clean room, the brand can compare its data on the users that ended up on the landing page for the new line of trainers, against the news website’s data on the users that were shown the advertisement. By comparing these two datasets, the brand can match the users that saw the advertisement with the users that ended up on the landing page and can, therefore, measure the campaign’s success.
- Audience overlap – Two or more parties can use data clean rooms to compare datasets for overlapping user data to develop co-marketing opportunities. For instance, a brand that sells desks might find that it has users in common with a brand that sells office chairs. Both companies could then work together by offering discounts or offers for users when they buy from both brands.
- Customer segmentation – Brands can use data clean rooms to better understand and segment their audiences, which in turn enables them to target their users with specific campaign objectives. For instance, an online music store such as HMV can put its customer data into a clean room which will, in turn, generate what it knows about HMV’s users. HMV can then use this information to segment its audiences based on certain differentiators and apply targeted campaigns about new releases, upcoming events, and more.
How do data clean rooms differentiate between each other?
There are two main types of data clean rooms that you need to know about:
- Those created by two or more partners: This is an independent type of clean room created by partnering companies in a neutral environment.
- Those offered by big tech platforms such as Google and Facebook: They are closed ecosystems where everything happens within that ecosystem. Their design makes them isolated, meaning they are incompatible with each other and any other data clean room platform.
Benefits and limitations of data clean rooms
- The use of data clean rooms allow advertisers to increase the size of their audiences fast by enabling them to match their data with publishers’ data
- Platforms are able to share data without actually ‘sharing data’ for matching
- The main blocker for using data clean rooms for first-party data comparison and exchange is that they add a significant data storage cost which increases proportionally to the amount of data that the publisher or advertiser upload. This essentially makes using them at scale anti-economical
- Interoperability is an issue because there are many different data clean rooms that exist yet they don’t integrate with each other. This means that their clients can only leverage the datasets of the companies that have integrated with that specific clean room.
- They lack cross-device capabilities – they predominantly compare ‘apple with apple’ i.e. emails to emails, MAIDs to MAIDs, etc. This means that publishers and advertisers using data clean rooms are unable to deploy their marketing strategies across devices. For most clean rooms, they become the ‘master’ of the data as they have access to the original dataset sitting within their infrastructure
What does the future hold for data clean rooms?
Working with data clean rooms as part of your addressability strategy offers a handful of different options depending on the requirements of your business. This includes a number of clear benefits, such as the ability for advertisers to build their audiences fast and the sharing of data without actually uncovering any personally identifiable information.
However, clean rooms have limitations, including their high costs and scale issues. Data clean rooms, in many cases, will act as an effective addition to an identity strategy, but shouldn’t be the one and only focus. Luckily, clean rooms can work in tandem with other solutions. In some cases, clean rooms can be used in conjunction with universal identifiers. The ID5 ID, for instance, can be used as a currency of comparison for the clean rooms.
Future-thinking players are exploring multiple approaches to make the most out of the opportunities provided by this new chapter of digital advertising. No doubt, we will see different solutions, including data clean rooms and universal identifiers, coexisting and used according to the needs of individual companies or campaigns.
Collaboration is key to a bright future for digital advertising, and data clean rooms are one important piece of the privacy-first puzzle that we as an industry are completing together.