As the world's digital transformation changes consumer expectations—and as third-party cookies are increasingly becoming a thing of the past—marketers are looking for a new way to make sense of customer behavior.
Many brands are looking to data clean rooms—but what are they, and are they right for you?
Well, let us explain.
A data clean room is a safe, pseudonymized space for known and prospective customer data. This allows marketers to analyze marketing and advertising data from many different sources in one, singular view while protecting the privacy of the data from each individual source.
This is most helpful in marketing and advertising contexts, where brands often have their own first-party data, data from partners and platforms (like Meta, Google, etc.) and permissioned or purchased data from third parties that they're trying to resolve across each data source. Data clean rooms allow brands to sync all of these data streams into one view of each person across these different contexts, increasing the value of the information they already own.
Data clean rooms are not new, but they are becoming increasingly popular as data privacy becomes more complex to navigate across geographies and platforms.
Many brands use data clean room providers for data collaboration. This is possible because of the stringent privacy controls built into the tech designed to protect all parties collaborating inside a data clean room. This tech allows brands to work with various data sets, including those from trusted partners, to augment and enrich data. This fills in data gaps on known and prospective customers, creating a richer understanding and analysis of each person.
These deeper insights allow marketers to engage customers based on their behavior across channels, and use first-party and third-party data to build audiences, activate media and provide measurement.
For example, a retailer who sells a CPG brand in their store may want to share data with that brand. Both companies serve the same customer, and sharing insights could improve how they market to specific individuals, creating a mutually beneficial reason to operate within a data clean room.
Walled gardens give brands the chance to reach a lot of prospects in the face of data deprecation. Platforms such as Facebook, Amazon, and Google offer data clean rooms to safely provide brands with ad performance data. These networks also tap into their huge consumer bases for audience targeting, using their own data — and that of the brand and its partners — to reach desired customers.
But with a greater focus on audience segments rather than individual behaviors, the insights gleaned from these platforms can be vague.
Inside a walled garden, it's hard to know whether a customer who falls into an ideal prospect category is actually in-market. If a walled garden only relies on weak identifiers—like a single email address—it can't see more holistic signals that indicate a person will be receptive to an ad.
This dovetails into a walled garden's measurement limitations.
Measurement is difficult within walled gardens because of the disparate nature of reporting and identity resolution across environments. Each walled garden differs in what measurement it offers, meaning metrics can vary across each platform and can be hard to compare. And it's extremely hard to track and understand users across these platforms.
Within a walled garden, a brand is limited to a per-platform snapshot and campaign-by-campaign performance. Without comparable results and clear identity, it's hard for marketers to assess the data needed to make critical decisions. Omnichannel visibility and collaboration require the full view of a customer's interactions.
Independent or third party data clean rooms are provided by technology vendors and are not tied to any single platform, making them channel-agnostic. In independent clean rooms, brands still get access to highly unique marketing data and don’t suffer from some of the platform restrictions encountered in walled gardens' clean room offerings. This is the most common choice for brands to adopt when deploying a data clean room.
These types of data clean rooms come in a wide variety of shapes and sizes. Some are point solutions — simple data clean rooms meant to provide straightforward service, whereas some are extensions of existing cloud data warehouses. Other data clean rooms have been acquired by larger companies or agencies and offer a wide range of use cases depending on the user.
But there's one thing that remains critical for any data clean room's success: access to data and identity. Most independent clean rooms come as an empty box that a brand is expected to fill. For brands without a lot of first-party data, it can be hard to find value in the clean room because they only have so muchhave limited data to analyze and glean insights from. They would need to rely heavily on partners and purchased third-party data to make the clean room investment valuable, thus tacking on costs to an already-expensive technology investment (not to mention the additional procurement involved).
And then there's the issue of data quality. Quality data is foundational for quality results. Brands often have disparate data that live in various systems and platforms, and a lot of that data is often inaccurate, duplicative or incomplete. Data clean rooms can deploy identity resolution solutions designed to unify, clean and organize data, but that often comes at an additional cost.
A data clean room equipped with pre-loaded data and identity enables brands to start activating their data on day one.
These data clean rooms are less common and are usually designed for large organizations and enterprises with customized data collaboration and collection needs. Typically, these clean rooms serve as a means for the organization to analyze its own data.
These types of data clean rooms offer maximum control for the owner and operator of the technology, but as a result, require a lot of time, energy and upkeep for said operator. These are often highly complex, require sophisticated expertise and incur significant costs.
Beyond data collaboration, a data clean room can help marketers do a number of things.
Today, brands typically work with limited information to deploy their marketing messages — even Amazon doesn't have every data point on an individual consumer to market to them effectively. Any brand is often limited by its own depth and breadth of data because it's specific to how the consumer has only interacted with them and they don't have a broader concept of how the consumer engages with other brands and on the open web. Without a broader picture of who they could potentially reach, there is only so much a brand can do with its own data.
Not all brands have a lot of data at their disposal. Let’s look at a real-world example using CPG brands.
Using a data clean room, brands can take their data and glean deeper insights using partner or third-party data. Once brands incorporate first-party, second-party and premium third-party data, like Epsilon’s proprietary data, they know who their customers are beyond their limited scope: what they buy (and why), where to find them and when to engage with them, driving retention and growth.
READ MORE: Understand your customer insights.
Increasing the quality and scope of existing data not only allows brands to talk to their known customers, but it also opens the door to all in-market customers.
Clean rooms help bridge that gap. A clean room that is equipped with identity and data enables brands to understand their current customers more deeply, build lookalike audiences based on their best customers and transform those unauthenticated customers into known ones.
At Epsilon, our proprietary data gives marketers insights into 250M+ unique U.S. individuals anchored in name and address to create audiences of those most likely to buy. It also helps identify who isn't in market, helping marketers spend media budgets more effectively.
READ MORE: Acquire new customers
With a complete, dynamic and persistent understanding of user behavior, preferences and demographics, you can personalize marketing campaigns across owned and paid channels, improving ad engagement and overall campaign performance.
And this matters: A survey from Epsilon on how consumers view personalized marketing and advertising shows that 76% of respondents said they view a brand negatively when they include inaccurate information about them in their marketing message. Even more interesting is that 91% said they see at least one irrelevant ad or marketing message a day.
The right data clean room enables brands to not only unify and expand their first-party data, but to gain a single, comprehensive view of their universe of potential buyers. And then use that view to craft a relevant message on the right channel at the right time. With person-first marketing, you’re able to understand, engage and learn from conversations with consumers on a 1:1 basis, across channels.
READ MORE: Personalize media across channels
Identity resolution and unified data in a data clean room not only provide insights before engaging a customer, but also measure campaign performance at the individual and aggregate levels.
Person-based marketing also enables person-based measurement. Brands can continually learn about their current customers and adjust acquisition strategies to find the best in-market people. Each time they deploy a campaign, the measurement helps affirm strong audience strategies and identify those that need adjustment.
READ MORE: Unify your brand and performance media
For most people, the word “AI” is associated with generative AI, which creates new content —such as images or videos — based on patterns learned from existing data. Most AI applications for martech/adtech require predictive AI. This type of AI looks for patterns across datasets to inform marketing strategies, autonomously deciding the best way to reach an outcome: who to talk to, where and when.
AI needs tons of data to learn and make the right predictions at scale. Brands typically have limited data on their own. Without a wide breadth and depth of data, AI won't deliver meaningful insights. If they don't deploy a data clean room equipped with data, they will need to augment their data by purchasing third-party datasets to achieve optimal AI results.
At Epsilon, we have 400TB of data that feeds our COREai. This allows our solutions to make 2+ billion model updates every minute and make 1,000+ trillion real-time decisions daily.
Scale alone isn't enough. AI can only learn from observation, meaning it's only as good as the data it's trained on. A recent Epsilon survey found that 49% of respondents are concerned that model accuracy is affecting efficacy.
A data clean room equipped with identity resolution can help prime a brand's data for AI engagement. Harmonized, cleansed, enhanced and connected data points provide a quality, accurate data foundation. A single identity spine creates stronger models for who their prospective customers might be.
Forrester Senior Analyst Stephanie Liu said data preparation is a critical step for brands looking to harness the power of AI. Without quality, accurate data, AI can't produce meaningful results. And as AI becomes more complex, quality control will become even more critical.
"AI is going to exacerbate those data issues," Liu said. "Just because the data is there doesn't mean its good data or the right data."
One quick-service restaurant client wanted to see a more persistent and connected view of their current guests and potential customers to drive more visits. Using an Epsilon clean room that housed multiple data sources—including their own—they gained a deeper understanding of their customers, including:
This, coupled with owned and competitive store visitation data, gave them insights into how to increase visits among existing guests and entice new guests to choose them over competitors, all through better advertising.
The results? They saw a 35% increase in in-store visits among the people they messaged, and identified four areas of growth for future messaging opportunities with newly acquired customers.
Epsilon Clean Room comes preloaded with data and identity, giving brands a foundational identity spine to bring first-party data together. We also offer proprietary audience data, giving brands a deeper view of their current customers and potential future customers.
But we go beyond simply having powerful tech. We offer pre-built predictive models and audiences for marketers to use, and access to audience strategists who can help with audience-first approaches and data strategies.
Learn more about Epsilon's Clean Room solution, how it works and what it can do for your business.
