By Lana Warner, Senior Director, Partnerships and Strategic Solutions, Lotame
Data clean rooms have been the topic of a lot of enthusiastic conversations for a while now, and today they’re certainly ready for the spotlight. The spotlight is a good place for them: Brands need a clear, up-close view of what clean rooms are capable of. This is especially timely, with data privacy regulations tightening up around the world. The promise of the data clean room is that it can provide a privacy-complaint, discreet environment for businesses to share their proprietary data. However, right now, brands and publishers find themselves lacking clarity about clean rooms’ best implementation for their own specific business needs.
The solution beyond the hype
When you break it down, the problem that the digital industry is trying to solve around third-party cookie deprecation is the problem of connectivity. Without data connectivity in the tech stack, brands’ core marketing initiatives are in jeopardy. Data clean rooms came on the scene at just the right time to be a contender in solving that problem. Brands would be wise to avoid snap judgments in how they invest in and implement clean rooms. For DCR (data clean room) success, marketers must consider several factors: their exact use case, their technical needs and resources (some DCRs require data scientists and more technical staff) and time to value (DCRs are not yet “plug and play,” and most do not offer real-time activation). They’ll also need to consider volume and types of customer data. To best navigate this landscape, it’s important to first understand the distinct categories of clean rooms, which have become clearer in recent months:
- Data Warehouses/Clouds. These are platforms that are data warehouses or clouds at their core, with clean room capabilities built on top. Snowflake, built on top of Amazon Web Services, is one familiar example. The business’s data lives in the warehouse, and it can be privately ported into the clean room for collaboration with partners, including for attribution and measurement. It’s possible to engineer highly customized solutions on these platforms, provided the participating business has strong engineering and data science capabilities of its own.
- Walled Gardens. These are the walled ecosystems of the largest tech businesses in the marketplace – think Google and Facebook (Meta). Marketers may find implementation to be relatively straightforward, although they could make customizations with the help of a data scientist from their own business. Keep in mind that when a marketer wants to bring their first-party data here, their data will need to be ported or integrated within the walled garden. These very large tech partners are not incentivized to send their data outside of their own ecosystem. Walled garden clean rooms are engineered for activation, suppression, measurement, and attribution. For example, we can imagine a marketer going to a Facebook clean room with DTC campaign data to better understand converting consumers, or to suppress current customers.
- Data Collaborations. In these environments, marketers bring in their own data, and combine it with a common identifier, which can range from HEMs and MAIDs, to Universal ID, powered by an identity spine to enable private, secure collaboration. This is particularly helpful for data analysis, enrichment, monetization, and activation. By contrast, a walled garden clean room will activate only within the walled garden itself. Marketers will also find data collaborations to be straightforward in implementation, as they’re built essentially for marketers to gain insights and activate seamlessly.
- These are environments where marketers can turn if they’re looking not only for data collaboration, but for neutrality in the digital ecosystem. The data is not portable and doesn’t need to be. As such, a query clean room functions as a bridge between different clean rooms and data warehouses. It’s owned by the business’s tech team, who will need to manage data orchestration across clean room participants. A more technically sophisticated marketer could do some of the crucial lifting, with training. One of the most important use cases here is understanding overlap. Query clean rooms will accept whatever common identifiers the brand and its partners prefer, and enable them to analyze overlap between collaborators.
Know the goal and take steps toward it
In pinpointing the category of data clean room that will best do the job at hand, marketers first need to understand where their data lives. Second, they’ll need to understand what their data represents. It matters, for example, whether or not their customer data is vastly authenticated, or whether the majority of their brand interactions are from unknown browser-based or in-app activities. Third, they’ll need to consider the identity solutions they’re working with.
Finally, a marketer will need a strong understanding of the use case. The first three factors combined – location of data, types of data, and identity solution – will inform how difficult or how easy it will be to address their use case, and in which clean room setting they can deliver the best results.
Marketers and brand leaders need to think about how little time they actually have left to prepare for third-party cookie deprecation, and about the level of urgency for making sure their data strategies are water-tight. Even brands with the most solid first-party data strategies may need clean rooms or “clean room-esque” capabilities to provide enrichment at scale, analyze overlap, enable data modeling beyond their existing consumer base, and power connectivity and activation. Brands can choose to pursue the opportunity clean rooms present to effectively conduct measurement and attribution after third-party cookie deprecation, or not.
Look to what the brand and the clean room can do together, and separately
Marketers also need to avoid expecting their clean room partners to provide benefits they’re not actually engineered to provide. They shouldn’t expect every clean room in the market will be able to compensate for any tech resources their own business doesn’t already have. Issues that brands find themselves considering with clean rooms include scale, available resources and support, overall cost, and the value they’ll get from that cost. Ethical data enrichment may demand a different tool – one that can ensure the data is both not visible one-to-one by all partners, and won’t enable re-identification back into a CDP, CRM, or any type of Customer 360.
Clean rooms, in summary, represent another tool in the marketers tool box, and are only as useful as the data that goes into them. Most cannot fix outdated email addresses, or a lack of email addresses that results in negligible overlap, or an unscalable amount of authenticated IDs for activation, or the business’s compliance with the ever-changing privacy legislation globally. Compliance with privacy law deserves specific mention here: Data clean rooms can help the business share auditable consent signals with other clean room participants. But they can’t make non-compliant data compliant. Gaining user consent is still the legal responsibility of the business that owns the data.
Ultimately, it’ll be up to marketers and brand leaders to determine what to look for in a clean room, and when the right time is to implement one. They’ll need to consider the use case and the available data, and consider whether one of their existing partners has clean room capabilities that meet their needs. An open mind is of the essence, as is a willingness to explore what’s working for other brands. There’s no one-size-fits-all solution.