AdCP aims to standardize how AI agents buy and sell ads. But can the industry align long enough to make it work—or will it be another acronym that fades?
The advertising industry loves a good standard. We’ve seen them come and go: OpenRTB to enable programmatic trading, MRC guidelines to define what’s viewable, ads.txt to clean up the supply chain, and the ARF to bring rigor to audience measurement. Even Google’s Privacy Sandbox, once hailed as the future of privacy-safe advertising, was quietly retired just last week.
Some standards worked brilliantly. Others created more problems than they solved.
Now, we have AdCP (Advertising Context Protocol) promising to standardize how AI agents communicate across the buy and sell sides. The first live agentic media buy happened last week. Industry reaction has ranged from “let’s wait-and-see how it develops” to “we’re automating insertion orders with the most advanced technology humanity has ever created.”
Both perspectives have merit. But the real question isn’t whether AdCP is revolutionary. It’s whether the conditions exist for it to succeed at scale. And who truly benefits if it does?
Lessons From Standards That Worked (and Didn’t)
OpenRTB succeeded for specific reasons. It assembled tech leaders from both supply and demand sides to develop an open protocol that solved a clear market need: enabling automated trading across platforms. Critically, it had broad buy-in from stakeholders of all sizes, not just the giants or start-ups who saw mutual benefit in standardization.
Compare that to ads.txt. Grand Theory: Verify Your Supply Chain Counterparties and Reduce Fraud. However, publishers immediately filled their ads.txt files with every possible partner, as limiting relationships meant limiting revenue. The standard was adopted, true. But it was essentially nullified by the very behavior it was designed to prevent. The standard failed, then, because publishers didn’t trust it.
Or consider seller-defined audiences (SDA). This was a fairly recent attempt (March 2021) to pass contextual audience data through the bid stream. It never gained traction because buyers didn’t trust sellers to define those audiences accurately. Also, there was no verification mechanism. The standard failed because incentives weren’t aligned.
Standards succeed when all parties trust the process and see tangible benefits; and standards fail when they’re gamed to preserve (perceived or real) individual advantage over collective progress.
What AdCP Needs to Flourish
For AdCP to move beyond demonstrations and pilot programs, five conditions must be met:
True tripartite adoption
To establish trust and promote uniformity, advertisers, agencies, and sellers must be equally represented in the development and evolution of standards. Currently, AdCP’s early coalition, initiated by Scope3, skews heavily toward supply-side platforms and ad tech companies. That’s not inherently problematic — someone has to start somewhere, right
However, sustainable success will require that demand-side platforms, holding companies, and major advertisers all actively shape the direction of the protocol. Every participant should ask: “Do I have a competitive advantage if everyone operates on the same standard?” If the answer is “I’ll need to customize it to fit my needs,” then the standard will likely fail in the face of yet another fragmented implementation.
Expansion beyond guaranteed buys
Starting with insertion orders and guaranteed inventory makes sense. It’s straightforward and high-value. However, the bulk of digital advertising occurs in performance-driven, programmatically traded environments. AdCP needs a clear roadmap for incorporating audience targeting, identity resolution, conversion tracking, real-time optimization, and measurement.
The protocol must quickly evolve to support the full spectrum of advertising workflows, not just the simplest ones. Otherwise, it remains a niche solution for premium direct buys, while the rest of the industry continues to operate through fragmented systems.
Integration with existing infrastructure
Here’s what many discussions about agentic advertising miss: AI agents don’t replace the platforms and systems that currently run advertising; they augment them. Every DSP, SSP, ad server, and measurement platform operates through APIs, proprietary interfaces that actually execute campaigns, serve ads, and report results.
AdCP provides a standardized language for agents to communicate, but those agents still need to connect to existing platform APIs to take action. The protocol only works at scale if there’s a clear path for integrating agent communication with the operational reality of today’s fragmented tech stack.
Privacy and data governance frameworks
First-party data no longer leaves enterprise firewalls. Companies won’t allow AI agents to freely access and activate proprietary customer data without robust controls, transparency, and a deployed architecture that keeps processing within their own environments.AdCP must address how agentic systems handle sensitive data and establish guardrails to prevent agents from making unauthorized decisions. It needs to determine how human oversight operates within automated workflows. Without addressing privacy and governance concerns, enterprise adoption stalls, regardless of how elegant the protocol may be.
Clear value proposition for each stakeholder
This is where most standards are either adopted or rejected. Publishers need one standard, not five competing implementations that force them to support multiple systems. Agencies need workflow automation that genuinely reduces operational overhead without eliminating their strategic value. Advertisers need better outcomes at lower cost with greater transparency.
If AdCP simply recreates the same margin pressures, hidden fees, and information asymmetry that plagues programmatic, it won’t matter how technically sophisticated it is. Nobody will adopt it at scale.
Who Wins, Who Loses, and Who Breaks Even
If AdCP attracts broad adoption, the winners are clear.
- Publishers need only one standardized protocol, rather than a myriad of customized integrations for every demand partner. Smaller publishers gain access to automated buying workflows that were previously available only to premium inventory, with dedicated sales teams.
- Agencies and trading desks that embrace agentic systems early gain efficiency advantages.
- Advertisers increase efficiency by accessing diverse inventory without incurring the costs of intermediary overhead.
The potential losers?
- Platforms whose competitive advantage relies on proprietary workflows and closed systems. If standardized agent communication reduces switching costs, market power shifts toward those who deliver the best outcomes, rather than those with the stickiest platform.
- Like any automation wave, some roles that primarily execute manual tasks will be displaced. The media buyers who survive and thrive will be the ones who understand the industry deeply enough to ask the right questions and interpret results; the ones who know math before picking up the calculator.
The Five-Year Outlook
Suppose AdCP achieves actual multi-stakeholder adoption and evolves beyond insertion orders. In that case, we’ll soon see parallel infrastructure and legacy programmatic systems running alongside agentic workflows, gradually shifting toward the latter as trust and capability expand.
If it fragments or gets gamed by individual platforms, creating incompatible implementations, we’ll have another acronym that promised transformation but delivered marginal improvement at best.
The deciding factor won’t be the technology. It will be whether the industry can align incentives long enough to build something that genuinely benefits everyone.
That’s a harder problem than building AI agents. But it’s the only one that actually matters.
