Privacy & Identity in Programmatic
The New Era of Responsible Advertising
The Privacy Revolution: Why Everything Changed
For over two decades, digital advertising relied on third-party cookies — small pieces of code dropped by websites to track users across the internet. This enabled behavioral targeting, frequency capping, and measurement. But it also meant users were tracked across thousands of sites without explicit consent.
Starting with GDPR in 2018 and accelerating with CCPA, iOS 14 changes, and Chrome's cookie deprecation, the industry is undergoing its biggest transformation since the invention of programmatic. The old model of cross-site tracking is ending. A new privacy-first identity ecosystem is emerging.
📅 Timeline: The Rise of Privacy Regulations
🌍 Major Privacy Regulations: GDPR, CCPA & Beyond
Enacted: May 2018
Scope: All companies processing data of EU residents
Consent: Explicit, opt-in required for processing
Key Rights: Right to access, right to be forgotten, data portability, right to object
Enforcement: Strong — major fines issued (Amazon €746M, Meta €1.2B)
Enacted: Jan 2020 / CPRA Jan 2023
Scope: Businesses with $25M+ revenue or 100,000+ consumers
Consent: Opt-out for data sales, opt-in for sensitive data
Key Rights: Right to know, right to delete, right to opt-out, right to correct
Enforcement: New dedicated agency (CPPA)
Other US State Privacy Laws
- Consent: Clear, affirmative action to opt in (GDPR) or opt out (CCPA)
- Right to Delete: Users can request deletion of their data
- Right to Access: Users can request what data is collected
- Data Minimization: Only collect data necessary for purpose
- Purpose Limitation: Data cannot be used for unrelated purposes
- Accountability: Organizations must demonstrate compliance
🔐 Consent Management & The Transparency & Consent Framework (TCF)
To operationalize privacy regulations, the industry developed Consent Management Platforms (CMPs) and the IAB Europe Transparency & Consent Framework (TCF). These enable publishers to capture and communicate user consent preferences across the programmatic supply chain.
How Consent Works in Programmatic
- User visits website: Consent Management Platform (CMP) displays banner asking for consent on data collection and ad targeting
- User chooses preferences: Accept all, reject all, or customize per purpose (e.g., accept analytics but reject personalized ads)
- Consent string generated: CMP creates a TC String (Base64-encoded) containing user's choices for each vendor and purpose
- Consent passed in bid request: The TC String is included in the OpenRTB bid request (in the
regs.ext.gdpranduser.ext.consentfields) - DSP enforces consent: DSP checks consent string against its vendor ID; only bids if user has consented to that purpose and vendor
Key TCF Concepts
BOEFEAyOEFEAyAHABDENA4CgAAAAAAAAAAThis string encodes: GDPR applies, user consented to purposes 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, but not to specific vendors. The DSP must decode this string to determine if it can bid.
🍪 The Cookieless Future: What's Replacing Third-Party Cookies?
With third-party cookies being deprecated across all major browsers (Safari, Firefox already block them; Chrome will phase out by end of 2025), the industry is building new identity and targeting infrastructure. No single solution will replace cookies — instead, a multi-layered approach is emerging.
Five Pillars of the Cookieless Future
Persistent identifiers based on hashed emails (deterministic) that work across browsers and devices with user consent. Users log in to publisher sites, enabling a consented identifier.
Targeting based on page content rather than user behavior. NLP-based analysis of keywords, sentiment, topics. Privacy-safe and gaining significant investment.
Secure environments where advertisers and publishers can match first-party data without sharing raw data. Enable audience overlap, measurement, and activation.
Publishers and advertisers investing in first-party data collection through logins, subscriptions, loyalty programs, and direct customer relationships. First-party data becomes the new currency.
Google's suite of APIs designed to enable interest-based advertising without cross-site tracking. Topics API, Protected Audience (FLEDGE), Attribution Reporting.
- Topics API: Browser observes user's browsing behavior and assigns topics (e.g., "sports," "travel") that are shared with advertisers. No cross-site tracking; only top-level interests.
- Protected Audience (formerly FLEDGE): Enables remarketing without cross-site tracking. Advertiser can add users to interest groups within browser; auctions happen locally.
- Attribution Reporting API: Enables conversion measurement without cross-site identifiers. Uses aggregated, noisy data to protect privacy.
🔑 Universal IDs: The Future of Identity
Universal IDs are deterministic identifiers based on hashed, encrypted emails. When a user logs into a publisher site, their email is hashed (converted to a unique string) and shared with participating partners. This creates a persistent identifier that works across sites, devices, and browsers — with user consent at the center.
Major Universal ID Solutions
Open-source universal ID led by The Trade Desk. Hashed email-based identifier. Supported by major DSPs, SSPs, and publishers. Open-source governance via Prebid.org. Dominant solution in North America.
LiveRamp's identity solution based on hashed email. Strong in data onboarding and clean rooms. Integrated with major DSPs and SSPs. Widely used for data collaboration.
Independent universal ID provider. Focus on privacy compliance and interoperability. Strong in Europe with TCF integration. Popular among European publishers.
European consortium-based ID solution. Focus on telco and media partnerships. Strong in Germany and other European markets. Alternative to US-dominated solutions.
🔒 Data Clean Rooms: Secure Data Collaboration
Data clean rooms are secure environments where two or more parties can match, analyze, and activate their first-party data without sharing raw data. They enable privacy-compliant audience overlap, measurement, and activation while maintaining data governance and privacy controls.
How Data Clean Rooms Work
- Data Onboarding: Publisher and advertiser upload their first-party data (hashed emails, user IDs) to the clean room.
- Secure Matching: Clean room matches users between the two datasets using encrypted identifiers. No raw data leaves each party's environment.
- Analysis & Activation: Parties can see aggregated insights (e.g., overlap size, performance metrics) without exposing individual user data. They can activate audiences (e.g., "users in overlap segment") to DSPs.
- Privacy Controls: Clean rooms enforce privacy thresholds (e.g., minimum cell size, noise injection) to prevent re-identification.
Major Data Clean Room Providers
- Audience Overlap: Discover how many customers are shared between publisher and advertiser
- Incrementality Measurement: Measure sales lift from campaigns using matched control/exposed groups
- Custom Audiences: Create and activate audiences based on combined data signals
- Cross-Channel Attribution: Understand how different channels contribute to conversions
- Frequency Management: Coordinate frequency across multiple buyers and publishers
🏠 First-Party Data: The New Currency
With third-party data becoming unavailable, first-party data — data collected directly from customers through owned channels — becomes the most valuable asset for both advertisers and publishers.
Types of First-Party Data
- Transaction Data: Purchase history, order value, product categories, frequency
- CRM Data: Customer profiles, loyalty program data, demographics, contact information
- Behavioral Data: Website visits, content consumption, search queries, app activity
- Engagement Data: Email opens, push notification responses, survey responses, customer service interactions
- Zero-Party Data: Explicitly shared preferences, interests, intentions (e.g., "I like running shoes")
First-Party Data Platforms
⚙️ How Identity & Privacy Flow in Programmatic
1. User visits site: User may have logged in (creating a universal ID) or be anonymous (cookie or device ID).
2. Consent captured: CMP captures consent choices. TC String generated and stored in user's browser.
3. Bid request created: SSP includes in bid request:
-
user.id: Cookie ID or universal ID-
user.ext.consent: TC String (GDPR)-
regs.ext.gdpr: GDPR applies flag-
regs.ext.us_privacy: CCPA opt-out string4. DSP receives bid request: Decodes consent string, checks vendor ID against user's preferences. If no consent for that vendor/purpose, DSP does not bid.
5. Bid response: If consent is valid, DSP may submit bid. Universal ID may be passed in
user.buyeruid.6. Creative served: Ad appears. Tracking pixels fire with appropriate privacy flags.
📚 Quick Reference: Privacy & Identity Essentials
EU privacy law. Explicit opt-in required. Fines up to €20M. Requires CMP and TC String.
California privacy. Opt-out for data sales. Right to delete. Fines $2,500-$7,500 per violation.
Hashed email-based identifier. UID2, RampID, ID5. Requires user login. Enables cross-site tracking with consent.
Secure data collaboration. Match first-party data without sharing raw data. Ads Data Hub, AMC, LiveRamp Safe Haven.
Google's replacement for cookies. Topics API, Protected Audience, Attribution Reporting.
Consent string passed in bid requests. Encodes user choices for vendors and purposes. Required for GDPR compliance.
Data collected directly from customers. Most valuable asset post-cookie. CDPs for activation.
Privacy-safe targeting based on page content. NLP-based analysis. Growing investment.