Discover how data clean rooms and Customer Data Platforms (CDPs) are reshaping modern marketing by preserving privacy while enabling data analysis in the face of stricter regulations and the decline of third-party cookies.
It was working well - advertising at scale on the internet was a solved problem, but then along came tighter privacy legislation, Apple announced changing the mobile tracking world with App Tracking Transparency, and soon 3rd-party cookies are going away. How are marketers supposed to do proper campaign measurement and optimization if they can’t gather data about customers and get acquainted with how they are responding to ads?
Commerce over the internet relies on the ability of organizations to track users across domains. However new privacy laws and the deprecation of third-party cookies have led to increased signal loss impacting marketers’ ability to efficiently plan and measure their campaigns. How do we maintain end-user privacy while preserving the ability to understand our customers and their journey?
Let’s first look at the current architectural landscape:
Companies use a Data Management Platform (DMP) to help assemble all of the data they gather (often third-party* data). DMPs rely heavily on third-party cookies for tracking users across multiple websites. Their ability to provide a holistic view of a user's journey and interests across the web is already being impacted by the depreciation of third-party cookies.
Since the announcement we’ve seen that many are moving to a Customer Data Platform (CDP), driven by 1st-party* data. First-party data in a CDP is arguably the most privacy-friendly way to operate. CDPs are designed to centralize first-party data from a variety of sources, including websites, CRM systems, mobile apps, and more. In a world moving away from third-party cookies, first-party data becomes even more valuable, and CDPs are well-suited to manage it.
However, we simply can’t learn everything we need to know about our customers through our direct interactions- some use cases require a view of the customer from an external perspective. This gap can be addressed by adding 2nd-party data* to the strategy… and this is where our initial problem statement enters the fray. How can we analyze data across organizational lines but maintain privacy?
*CALLOUT: For a refresher on first, second, and third-party classifications of data, check out this blog.
The problem is that we need to share data without exposing the customer’s identity; but how? So engineers did what engineers always do - they created a way to meet the need! The data clean room has emerged as the answer to this riddle.
A data clean room is a secure environment where advertisers, agencies, and publishers safely analyze and share data without revealing any personally identifiable information (PII), allowing them to adhere to privacy laws and regulations. They incorporate Privacy Enhancing Technologies (PETs) like encryption and differential privacy to protect individual identities and data misuse. Additionally, DCRs enforce strict access and privacy controls, ensuring users only access necessary data and resources for specific tasks. Data clean rooms are typically used by businesses to perform cross-brand marketing collaboration, measure the effectiveness of campaigns on advertising platforms, and gain insights into customer behaviors and attributes.
By no means is this list comprehensive, but here are some common uses and benefits of clean rooms that organizations employ:
Campaign Measurement, Planning & Attribution: Clean rooms enable advertisers to gauge the efficacy of their campaigns. They can determine the overlap between conversions on their site and impressions on publisher and walled garden platforms, to drive a holistic understanding of audience exposure, impact, and ad performance. This leads to a more accurate measurement of return on advertising spend (ROAS). Publishers can allow advertisers to answer questions like, how will a certain audience ad campaign perform on this platform, or what was the performance of an executed campaign?
Profile Enrichment: By collaborating in a clean room environment, companies can augment their existing customer data with information sourced from partner collaboration. For instance, a credit card company could partner with a data analytics firm to gain insights into user shopping behaviors, thereby refining their targeted marketing campaigns.
First-party Data Partnerships: Data clean rooms catalyze collaborations, enabling brands to delve into new segments and uncover opportunities for brand growth. Clean rooms empower marketers to create precise look-alike models, fostering segmentation based on consumer behavior signals from partners. Two or more brands can assess the performance of their joint campaigns via 1P customer transactional data (LTV, purchase history, etc.) Imagine an airline partnering with hotels, or CPGs partnering with retail companies. All of these brands can leverage insights from collaborations to power a connected experience across their platforms and issue personalized offers.
The relationship between a CDP and a data clean room is symbiotic. A CDP is essential for collecting and feeding organized, high-quality first-party data into the data clean room, forming the basis for deeper, multi-party analysis. In turn, the aggregated insights derived from the data clean room can be channeled back into the CDP, enriching customer profiles and enhancing data-driven activation strategies.
In essence, while a CDP organizes and makes available the necessary first-party data, the data clean room extends the analytical landscape by enabling privacy-compliant, collaborative analysis with broader data sets within publisher platforms, walled gardens, and industry-adjacent brands.
There are several types of data clean rooms with different user experiences. Independent data clean rooms like Habu and Infosum provide a neutral space for multiple parties to collaborate, and often include enrichment capabilities. Walled-Garden clean rooms such as Amazon Marketing Cloud offer a space for advertisers to match their data against their ad exposure logs, with standard reporting.
There is also an emergence of data onboarding vendors providing clean room collaboration alongside identity resolution and access to data marketplaces. All of the leading Data Warehouse and Data Lake vendors provide data clean room service products to allow brands and publishers to build their own data clean rooms. While the tooling and user experience may differ based on the type of data clean room, all clean room share the below principles in how they work:
Data Ingestion
First-party data is sent or securely shared to a clean room from both parties. Many clean rooms onboard via data warehouses while some also offer ingestion APIs.
Data Privacy and Matching
Matching data sets between the parties requires a common identifier. Collaborators can leverage hashed PII and/or durable cookieless IDs such as RampID or UID2 to do the matching. With the current clean room match rate around 50%, collaborators often look towards identity enrichment providers to help solve the identity interoperability challenge and maximize their collaboration ROI.
Data clean rooms employ a variety of technologies and methodologies to de-identify and secure the data they handle, ensuring that PII is removed or encrypted. Some of the key cryptographic technologies include anonymization, hashing, differential privacy, and homomorphic encryption all aimed at ensuring data cannot be re-identified.
Analytics and Enrichment
Data is analyzed at this stage for measurement, attribution, overlaps, and scoring. Clean rooms provide a variety of experiences including SQL based analysis and standard reports. Some clean rooms have AI capabilities to help marketers attain suggested audiences based on collaborated data or LLMs to allow marketers to enter business questions reducing SQL knowledge requirements.
Activation
In a typical activation layer, marketers can execute campaigns based on the analysis and aggregated match outcomes. There are strict parameters on what data can leave the clean room. The major wall-garden clean rooms tend to provide easy activation into their platforms from their clean room collaboration products. To increase the scope of activation, marketers are looking for more one-to-many programmatic activation platforms to maximize their reach.
What value does a CDP add to your clean room implementation?
Data organization is still vital. As the industry moves towards data clean rooms as the preferred way to collaborate, brands, small and large, will require substantial first-party data to avoid being excluded from the ecosystem. Segment CDP builds complete, privacy-compliant first-party data profiles to feed the data clean room. These profiles contain all customer traits, audience membership, and advanced calculations available as data points for each customer. The cleanliness of your profile information allows you to easily ensure you’re compliant, as trying to manage that cleanliness manually is a tremendous lift for data teams.
Identity interoperability is a hard problem. Twilio Segment sets the business up with the proper first-party-based identity graph ready for onboarding with industry-leading onboarding platforms like LiveRamp.
Leverage for the AI Revolution. Segment provides an easy to use audience building and advanced AI/ML capabilities to enhance the data that a brand brings into a clean room. AI success absolutely, unequivocally requires a solid data foundation, or it will crumble.
Build once, use everywhere. Many businesses will find themselves participating in multiple data clean room products and collaboration spaces. Segment will provide a single upstream source of first-party data to all of these collaborations to ensure a consistent foundation for measurement and enrichment. Segment is integrated with all major data warehouse-based clean rooms, and with many other types of clean room solutions such as Google Ads Hub, LiveRamp SafeHaven, Habu. It also reduces the time-to-value to build your own Data clean room.
Organizing your inputs is only half the story. Segment can extract insights resulting from clean room based look-alike modeling and other insights from the warehouse. This golden profile of a customer from all of these data sources can then be shared openly with the rest of your stack, ensuring that every tool has the data it needs to maximize its effectiveness.
Centering your post-DMP data approach around a customer data platform is essential for a robust first-party data strategy. The CDP’s consolidation of all customer interactions across the entire journey builds a comprehensive view that is critical in providing large-scale personalized and pertinent experiences for customers.
Since we can no longer look to third-party data to provide insights and reliably fuel marketing strategies, we turn to the data clean room as a way to provide second-party data - which can give us powerful insights about customer behaviors and preferences, but while preserving privacy.
Twilio Segment empowers clean room projects with a strong first-party data foundation and delivery mechanisms to send that data to all clean room ingestion points to maximize collaboration ROI.
Our annual look at how attitudes, preferences, and experiences with personalization have evolved over the past year.