Cookieless & data clean rooms: real-world application

CORE CONCEPT: WHAT IS ACTUALLY CHANGING AND WHY IT MATTERS

Third-party cookies are small tracking files that advertisers have used for 25 years to follow users across websites, build behavioral profiles, and retarget them. When a browser blocks those cookies, the entire infrastructure of cross-site audience tracking collapses. You cannot see if the person who visited your product page also saw your ad on a news site. You cannot build lookalike audiences from browsing behavior. You cannot frequency cap across publishers.

A data clean room is a secure, privacy-safe environment where two or more companies can bring their first-party datafirst-party dataData collected directly from your own customers and prospects through your own channels: your most reliable and privacy-compliant source.View full definition → together, run analyses, and extract insights without either party ever seeing the raw records of the other. Think of it as a locked room where you slide your customer list under one door, a publisher slides their audience data under another door, and a neutral system tells you the overlap without handing you the actual names or emails. Google Ads Data Hub, Amazon Marketing Cloud, and LiveRamp Clean Room are the three most used in practice right now.

First-party dataFirst-party dataData collected directly from your own customers and prospects through your own channels: your most reliable and privacy-compliant source.View full definition → means data you collected directly from your own customers through your own channels: purchase history, email signups, loyalty programs, app behavior, CRMCRMCustomer Relationship Management: software and strategy to manage and analyse customer interactions throughout their lifecycle.View full definition → records. This is the asset that replaces the cookie. If you own it, you have targeting. If you do not, you are renting someone else's audience at a premium forever.

KEY SUB-CONCEPTS WITH REAL EXAMPLES

1. IDENTITY RESOLUTION WITHOUT COOKIES

Identity resolution is the process of connecting fragmented signals (a device ID here, an email address there, a hashed login somewhere else) into a single unified customer profile. LiveRamp's identity graph connects over 500 million pseudonymous profiles across devices and environments without relying on third-party cookies. Procter and Gamble uses LiveRamp's RampID to match their loyalty data against publisher audiences and run targeted campaigns on premium inventory without cookies. The result reported by LiveRamp in their 2022 case documentation showed P&G achieving match rates above 60% in cookieless environments, compared to near-zero in unresolved third-party contexts.

2. DATA CLEAN ROOMS IN MEDIA BUYING

When you activate a clean room partnership with a publisher like NBCUniversal (their platform is called One Platform), you upload a hashed version of your CRMCRMCustomer Relationship Management: software and strategy to manage and analyse customer interactions throughout their lifecycle.View full definition → emails. NBCUniversal matches that against their authenticated viewership. You get reachreachThe number of unique people exposed to your message in a given period. Unlike impressions, reach counts each person once, no matter how often they see it.View full definition → and frequency data, conversion lift analysis, and audience overlap reports. You never see their viewer records. They never see your customer list. This is how Expedia Group ran its streaming TV attributionattributionA framework for assigning credit to the touchpoints that contributed to a conversion, so you can measure which channels and interactions actually drive results.View full definition → strategy in 2023, using NBCUniversal's clean room to prove that connected TV exposure was driving actual hotel bookings in their loyalty database, with a reported 22% lift in booking intent among exposed audiences.

3. FIRST-PARTY DATAFIRST-PARTY DATAData collected directly from your own customers and prospects through your own channels: your most reliable and privacy-compliant source.View full definition → ACTIVATION THROUGH RETAIL MEDIA

Retail media networks are exploding because they are built on first-party purchase data by design. Amazon Advertising, Walmart Connect, and Kroger Precision Marketing all operate on the principle that they know what their customers bought, not just what they browsed. Kraft Heinz shifted a portion of their shopper marketing budget to Kroger Precision Marketing and used Kroger's 84.51 analytics division to match product purchase data against household profiles. The reported outcome was a 30% reduction in wasted impressionsimpressionsThe total number of times an ad or piece of content is displayed, regardless of clicks. Each display counts as one impression, even to the same person.View full definition → compared to traditional CPG media buying, because the targeting was grounded in actual purchase behavior rather than inferred intent from cookies.

4. SERVER-SIDE TAGGING AS OPERATIONAL INFRASTRUCTURE

Server-side tagging moves your tracking logic off the user's browser and onto your own server. Instead of a dozen JavaScript tags firing in the browser and being blocked by ad blockers or privacy settings, your server sends clean, verified signals directly to Google Analytics 4, Meta's Conversions APIAPIApplication Programming Interface: a standardised interface that lets applications communicate and exchange data without knowing each other's internal workings.View full definition →, or TikTok's Events APIAPIApplication Programming Interface: a standardised interface that lets applications communicate and exchange data without knowing each other's internal workings.View full definition →. Gymshark implemented server-side tagging via Google Tag Manager server containers in 2022. They recovered approximately 15 to 20% of previously lost conversion signals according to their reported implementation results, which directly improved their Meta ROASROASReturn on Ad Spend (ROAS) measures the revenue generated for every unit of currency spent on advertising, calculated as revenue divided by ad cost.View full definition → by giving the algorithm better training data.

REAL-WORLD CASES WITH RESULTS

CASE 1: DISNEY AND THE CLEAN ROOM PREMIUM

Disney built its own clean room called Disney's Advertising Clean Room, launched in 2022. Advertisers like L'Oreal and Unilever use it to match their CRMCRMCustomer Relationship Management: software and strategy to manage and analyse customer interactions throughout their lifecycle.View full definition → data against Disney's authenticated streaming audience across Disney+, Hulu, and ESPN+. Disney reported that advertisers using the clean room saw audience match rates of 75% or higher because Disney requires authenticated logins across all its platforms, meaning nearly every viewer is a known identity. L'Oreal used this to suppress existing customers from acquisition campaigns and saw a measurable reduction in wasted spend in their beauty category campaigns during Q4 2022.

CASE 2: HEINEKEN AND GOOGLE ADS DATA HUB

Heineken used Google Ads Data Hub to analyze the overlap between YouTube ad exposure and in-store purchase data via their retail partners in the Netherlands. The Ads Data Hub environment allowed them to run custom SQLSQLSales Qualified Lead: a prospect the sales team has validated as ready for direct outreach and a proposal, having passed clear qualification criteria.View full definition → queries on their campaign data without accessing individual user records. The outcome was a reallocation of their video budget away from broad reachreachThe number of unique people exposed to your message in a given period. Unlike impressions, reach counts each person once, no matter how often they see it.View full definition → formats toward targeted TrueView placements, with a 17% improvement in incremental reachreachThe number of unique people exposed to your message in a given period. Unlike impressions, reach counts each person once, no matter how often they see it.View full definition → among verified beer purchasers reported through their agency Publicis.

CASE 3: THE HOME DEPOT AND FIRST-PARTY DATAFIRST-PARTY DATAData collected directly from your own customers and prospects through your own channels: your most reliable and privacy-compliant source.View full definition → MONETIZATION

The Home Depot launched its retail media network, Orange Apron Media, in 2023. They built it entirely on their Pro loyalty program data and in-store purchase history. Suppliers like Milwaukee Tool and Makita now buy media on The Home Depot's owned channelsowned channelsMedia channels a company owns and controls directly, such as its website, blog, newsletter, social accounts and mobile app. No per-use payment to a publisher is required.View full definition → using purchase-based targeting. This is not just a revenue stream. It is proof that first-party datafirst-party dataData collected directly from your own customers and prospects through your own channels: your most reliable and privacy-compliant source.View full definition → infrastructure pays for itself by becoming a media asset you can sell access to, not just use internally.

CMO ACTION ITEMS

• Audit your first-party datafirst-party dataData collected directly from your own customers and prospects through your own channels: your most reliable and privacy-compliant source.View full definition → coverage this quarter: mapmapUsing software to automate repetitive marketing tasks and campaigns, enabling personalisation at scale across channels like email, web, and social.View full definition → every touchpoint where you collect authenticated user data (email, loyalty, app login, CRMCRMCustomer Relationship Management: software and strategy to manage and analyse customer interactions throughout their lifecycle.View full definition →) and calculate what percentage of your media spend currently has a matchable first-party signal attached to it. If that number is below 50%, you have a structural problem before you have a targeting problem.
• Initiate one clean room pilot with a major publisher or platform you already spend with: pick Google Ads Data Hub if you run YouTube, or Amazon Marketing Cloud if you run sponsored ads. Run a 90-day conversion path analysis. The goal is not a perfect result. The goal is to understand your attributionattributionA framework for assigning credit to the touchpoints that contributed to a conversion, so you can measure which channels and interactions actually drive results.View full definition → gaps before your agency reports tell you what they want you to believe.
• Implement Conversions APIAPIApplication Programming Interface: a standardised interface that lets applications communicate and exchange data without knowing each other's internal workings.View full definition → or server-side tagging on your top two conversion events: work with your engineering team to push server-side signals to Meta and Google. This is not optional for performance marketing in 2024. Every unconverged signal is a lost optimization cycle for the algorithm.

COMMON MISTAKES THAT KILL RESULTS

• Treating clean rooms as a vendor relationship instead of a data strategy: most CMOs sign a clean room contract because their agency recommended it or a publisher pitched it. Then the clean room sits unused because no one on the team knows how to write the queries or interpret the outputs. A clean room without an analytics team that owns it is an expensive license you are not using. Hire or assign a dedicated person before you sign.
• Confusing data enrichment with first-party datafirst-party dataData collected directly from your own customers and prospects through your own channels: your most reliable and privacy-compliant source.View full definition →: buying third-party audience segmentssegmentsDividing a market into distinct groups of customers who share similar needs, characteristics or behaviours, so each group can be served with a tailored approach.View full definition → from a data broker like Acxiom or Experian and layering them onto your CRMCRMCustomer Relationship Management: software and strategy to manage and analyse customer interactions throughout their lifecycle.View full definition → is not first-party datafirst-party dataData collected directly from your own customers and prospects through your own channels: your most reliable and privacy-compliant source.View full definition → strategy. It is still renting someone else's inference. The distinction matters legally under GDPR and CCPA and commercially because those segmentssegmentsDividing a market into distinct groups of customers who share similar needs, characteristics or behaviours, so each group can be served with a tailored approach.View full definition → degrade in quality as signal loss accelerates. Build the consent-based collection infrastructure directly.
• Skipping the consent and governance layer: clean rooms do not automatically make your data practices compliant. If you collected emails without explicit consent for advertising use, running them through a clean room does not fix the legal exposure. Unilever found this out the hard way in several European markets where their clean room activations were audited under GDPR's purpose limitation principle. Get your privacy counsel involved before the data science team, not after.