Understanding the Complicated World of Advertising Technology (AdTech)
This guide covers the full arc of digital advertising technology: key terminology, a brief industry history, the core platforms and processes that power programmatic advertising today, and the major trends and challenges shaping its near-term future.
Key Terms
Online advertising vs. digital advertising vs. programmatic advertising
Online advertising usually refers to ads displayed within web browsers.
Digital advertising is the broader category — any advertising delivered through a digital channel.
Programmatic advertising is the automation of digital advertising buying and selling through software and data.
AdTech vs. MarTech
AdTech platforms are software tools used to buy, sell, and measure digital advertising.
MarTech platforms are software tools used to create and manage broader digital marketing activities.
First-party vs. third-party data and cookies
First-party data (and cookies) is data collected directly from a customer or visitor by the property they're interacting with.
Third-party data (and cookies) is data collected indirectly — gathered by entities other than the site the user is actually visiting.
A Brief History of Digital Advertising
Three decades of online advertising can be summarized through three defining themes:
- 1990s: Innovation
- 2000s: Acquisitions
- 2010s: Privacy
The 1990s — Foundations
- 1994: Web cookies are invented.
- 1994: The first banner ads appear. The first ever online ad ran on HotWired (now Wired.com) on October 27, 1994, placed by AT&T.
- 1995: The first ad server is founded.
- 1996: Kevin O'Connor, Dwight Merriman, and Fergus O'Daily found DoubleClick.
- 1996: Yahoo! begins displaying search ads on its search engine.
The 2000s — Consolidation Through Acquisition
- 2000: Google launches AdWords (now Google Ads).
- 2002: Applied Semantics creates contextual advertising technology, which later forms the basis of AdSense.
- 2007: Mobile ad networks — including AdMob — emerge and begin selling ad space on mobile devices.
- 2007: Google acquires DoubleClick for $3.1 billion.
- 2007: Microsoft buys AdECN for a reported $50–75 million.
- 2007: Yahoo! purchases RightMedia for $700 million.
The 2010s to Now — Privacy Takes Centre Stage
- 2016: The European Union releases the General Data Protection Regulation (GDPR).
- 2017: Apple Safari releases its Intelligent Tracking Protection (ITP) feature.
- 2018: Mozilla Firefox releases its Enhanced Tracking Protection (ETP) feature.
- 2018: California introduces the California Consumer Privacy Act (CCPA).
- 2020: Google announces it will shut off third-party cookie support by 2022 (later revised to 2023).
- 2020: Apple announces significant new privacy changes to its mobile operating systems.
The Main Platforms, Processes, and Players
Advertisers and Publishers
Advertisers are brands or companies (including agencies acting on their behalf) that want to get their products or services in front of a target audience — to build brand awareness, develop loyalty, and drive sales.
Publishers are companies that produce content that attracts an audience and monetize that audience through advertising.
How the Programmatic Advertising Industry Is Structured
Programmatic advertising operates across a demand side and a supply side:
Demand / buy side
- Advertiser's ad server
- Demand-side platform (DSP)
- Data management platform (DMP)
- Customer data platform (CDP)
Supply / sell side
- Publisher's ad server
- Supply-side platform (SSP)
Intermediaries
- Ad networks
- Ad exchanges
The major players in the industry fall broadly into three categories:
- Agency holding companies
- Walled gardens (Google, Meta, Amazon, Apple)
- Independent AdTech and data companies
Major Trends, Challenges, and Opportunities
Privacy
Privacy Laws
Governments around the world are introducing and updating privacy and data protection regulations. Most require companies to obtain user consent before collecting data and give users greater control over how their information is used.
The EU's GDPR came into force on May 25, 2018, and is widely regarded as the catalyst for the wave of new and updated privacy laws that followed globally. The CCPA is California's equivalent, also introduced in 2018. Many other US states have since introduced or are developing their own frameworks.
Web Browsers and Third-Party Cookies
Apple's Safari and Mozilla's Firefox block third-party cookies by default and have added additional privacy-protection features (ITP and ETP respectively). Google Chrome will deprecate support for third-party cookies, with the timeline now pointing to 2023.
First-party cookies are created by the website the user is currently visiting.
Third-party cookies are created by other companies — not by the site the user is on — and are used for cross-site identification (also called cross-site tracking).
Once a user can be identified across different websites using third-party cookies, the following processes become possible:
- Audience targeting and retargeting
- Frequency capping
- Measurement and attribution
Mobile IDs (IDFA)
Apple's changes to its mobile identifier for advertisers (IDFA) impose similar constraints in the mobile app environment, requiring explicit opt-in consent from users before apps can access the identifier.
Privacy Solutions
Google Chrome's Privacy Sandbox
The Privacy Sandbox is a set of open standards designed to replace the functions currently carried out by third-party cookies — including interest-based audience targeting, retargeting, measurement, and attribution — without enabling 1-to-1 user identification. Instead, everything would be processed in aggregate form to strengthen user privacy.
Key proposed APIs include Topics API, TURTLEDOVE, FLEDGE, and others. The plan is to launch alongside Chrome's deprecation of third-party cookies in 2023; Chrome is unlikely to remove third-party cookies until Privacy Sandbox is ready.
Current state (with third-party cookies):
| Function | Mechanism |
|---|---|
| Audience targeting | Third-party cookie (3PC) |
| Frequency capping | Third-party cookie (3PC) |
| Retargeting | Third-party cookie (3PC) |
| Measurement & attribution | Third-party cookie (3PC) |
Privacy Sandbox replacement model:
| Function | Mechanism |
|---|---|
| Contextual targeting | Retained |
| Audience targeting | Topics API, FLEDGE, etc. |
| Retargeting | TURTLEDOVE / FLEDGE |
| Measurement & attribution | Privacy Sandbox APIs |
ID Solutions
Several industry ID solutions have emerged as alternatives to third-party cookies for cross-site identification. The main ones include:
- The Trade Desk's Unified ID 2.0
- LiveRamp's ATS (Authenticated Traffic Solution)
- Tapad's Switchboard
- Neustar's Fabric ID
- Nielsen's Sync
- Permutive
- ID5
Most of these solutions rely on publishers collecting authenticated identifiers — typically email addresses or phone numbers — from visitors before an ID can be created. Even where these solutions work in a conceptually similar way to third-party cookies, they won't achieve the same scale, since they depend on user authentication that isn't universal.
ID Graphs and Data Clean Rooms
Identity graphs and data clean rooms are additional privacy-preserving mechanisms that allow data collaboration and audience matching without exposing raw user-level data.
Ad Fraud
Ad fraud remains one of the industry's most significant and persistent problems:
- Juniper Research estimates that ad fraud will cost the online advertising industry $100 billion by 2023.
- The WFA (World Federation of Advertisers) predicts ad fraud will become the biggest market for organized crime by 2025, worth $50 billion.
- Human (formerly White Ops) and the Association of National Advertisers reported that $6.5 billion was lost to ad fraud in 2017, down from $7.2 billion in 2016.
Antitrust Investigations
Governments in both the US and Europe have launched investigations into the dominant technology platforms — Google, Meta (Facebook), Amazon, and Apple — examining whether their practices in advertising and broader technology markets constitute anticompetitive behaviour. Both Google's and Meta's advertising businesses have come under particular scrutiny from regulators on both sides of the Atlantic.
Blockchain
Opinion in the industry is sharply divided on whether blockchain can meaningfully address programmatic advertising's challenges.
Arguments in favour:
- Could make contracts and payments faster.
- Could unify measurement.
- Could increase transparency across the supply chain.
Arguments against:
- Blockchain's transaction speed is too slow for the high-frequency demands of programmatic advertising.
- It is not optimized for high-transaction systems and is therefore not well-suited to real-time bidding environments.
Problems blockchain is proposed to solve:
- Transparency in the media supply chain (pricing, commissions, measurement).
- Ad fraud.
- User privacy.
- Payments and reconciliation.
- User incentives — for example, the Brave browser's Basic Attention Token (BAT) model, which compensates users directly for viewing ads.
Artificial Intelligence and Machine Learning
AI and ML have more genuinely applicable use cases in programmatic advertising than blockchain does.
Key definitions:
- Artificial intelligence (AI): Technology that aims to carry out activities currently performed by humans.
- Machine learning (ML): A subset of AI that allows a machine to automatically learn from data without requiring explicit reprogramming.
- Deep learning: A subset of machine learning that generates insights from data, focusing on more specific aspects of the learning process.
Benefits of ML and AI in programmatic advertising:
- Improved campaign performance (clicks and conversions)
- Optimized media spend
- More valuable audience construction
Note: ML and AI require large amounts of data to deliver meaningful benefits.
Common use cases:
- Personalization
- Content matching and contextual targeting
- Budget and bidding strategy optimization
- Predictive analysis
Connected TV (CTV) and Over-the-Top (OTT) Advertising
Connected TV (CTV)
Connected TV refers to devices that connect to the Internet and allow viewers to watch video content from streaming services. CTV devices include smart TVs, game consoles, and plug-in streaming devices.
Importantly, the IAB does not classify desktops, laptops, or mobile devices (smartphones and tablets) as connected TV, even though those devices are commonly used to stream video content.
Over-the-Top (OTT)
Over-the-top (OTT) refers to video content delivered via the Internet rather than through traditional broadcast networks (also known as multiple-system operators, or MSOs). The "over-the-top" name reflects the fact that these services are delivered over the Internet, bypassing legacy cable and satellite infrastructure.
Popular OTT services include Netflix, Hulu, Roku, Amazon Prime, and HBO Go.
The Metaverse — Augmented Reality (AR) and Virtual Reality (VR)
Augmented reality and virtual reality environments represent an emerging frontier for digital advertising. As the metaverse concept develops, the industry is watching to understand how advertising formats and targeting mechanisms will need to evolve for immersive, spatial environments.
Source: IAB
Programmatic advertising continues to evolve rapidly across all of these dimensions. The practical implication for anyone operating in this space is that no single challenge — privacy, fraud, measurement, or emerging channels — exists in isolation. The solutions being developed across cookies, identity, AI, and CTV advertising are interconnected, and understanding how the pieces fit together is essential for navigating what comes next.