Your Martech Stack Is Becoming Someone Else's Architecture

The martech stack is likely dissolving into company-wide data and AI architecture, but that’s a forecast, not today’s reality. The Monday-morning move: find out where your customer data actually lives, who owns it, and whether marketing has a seat at the architecture table.

In his May 14, 2026, newsletter article , Scott Brinker (Independent Analyst & Advisor, chiefmartec.com) is right about the direction. Three to five years from now, “the martech stack” as a marketing-owned system probably won’t exist in its current form. Marketing will be one tenant on a company-wide data and AI architecture. But here’s what most of the commentary is skipping: the gap between that vision and where marketing teams sit today. And the work to close it doesn’t start with architecture. It starts with a question most CMOs can’t answer yet.

What Brinker Is Saying

Scott argues that marketing is becoming an architectural discipline. His frame: the martech stack is raw materials, marketing architecture is the finished structure. He borrows a 2,000-year-old framework from Vitruvius to test marketing systems on three dimensions: utility (does it serve outcomes?), durability (can it adapt without breaking?), and beauty (is the experience distinctively yours?). He layers in a “context stack” of six pace layers, from slow-moving market strategy down to real-time moment decisions (1. Brinker, 2026).

Smart framework. Also a forecast, not a field report. And the distinction matters more than the commentary acknowledges.

Where Most Teams Are Now

Spencer Stuart surveyed 100+ CMOs: 51% rate their ability to use AI for marketing objectives as merely average. The CTO or CIO leads AI strategy in 44% of organizations; the CMO leads it in 32%. The top barrier to scaling AI across marketing: technology integration, martech, adtech, data, and legacy systems tangled together (2. Spencer Stuart, 2025).

In most organizations, marketing doesn’t own the architecture Brinker describes. It doesn’t control the data that feeds it. And it isn’t leading the AI strategy that will reshape it.

Then there’s the golden record problem. Scott’s framework assumes something like a unified customer data foundation that marketing can build on. The industry has been promising that single view of the customer for 15 years. Most enterprises still can’t answer a straightforward question: where does our customer data actually live, who owns it, and what’s marketing’s authority over it? That is a huge governance gap, and no platform purchase resolves it.

Even with 69% of organizations now running generative AI, 57% of data leaders still cite data reliability as a top barrier to getting AI projects into production (5. Informatica/Wakefield Research, 2026). The tools keep advancing. The data foundation doesn’t keep pace.

The CDP was supposed to be the answer. For many organizations, it hasn’t been. Not because CDP technology failed, but because they bought a tool to fix a problem that was never technical. Warehouse-native approaches are gaining traction precisely because they force the data ownership question earlier. But swapping one platform category for another doesn’t fix governance.

One CMO in the Spencer Stuart survey said it directly: AI depends on its data foundation, and the CMO is often left figuring that out alone (2. Spencer Stuart, 2025).

The Forecast Is Directional. The Work Is Immediate.

Scott is describing where this is heading, not where most organizations are. But directional signals are already visible. Marketing ops is converging with revenue ops in practice: shared definitions, joint dashboards, combined tech ownership across marketing, sales, and customer success (3. Ciberspring, 2026). The responsibilities have merged even when the reporting lines haven’t. The best marketing organizations treat marketing data with the same rigor and governance as they apply to financial or HR data assets (4. Transparent Partners, 2025). Most aren’t close to that standard yet.

So what do you do with a forecast you believe is directionally correct but that describes a world you don’t live in yet?

You start where you are.

What to Do Monday

If you’re a CMO: Get a one-page answer to three questions. Where does our customer data actually live? Who owns it? What is marketing’s authority over it? If nobody in your organization can answer those cleanly, you’ve found your starting problem. The CTO or CIO is already leading AI strategy in nearly half of all organizations (2. Spencer Stuart, 2025). That’s a vacuum being filled, not a power grab. If marketing doesn’t show up to the architecture conversation with informed questions, the architecture gets built around engineering’s priorities. You don’t need to become an engineer. You need to know who’s making the data decisions that will shape your future stack, whether you’re in the room or not.

If you’re an ops pro: Pick one capability area (lead routing or audience activation) and trace it end to end. Where’s the data coming from? Who decides the logic? What would change if an agent made that decision instead of a rule? That gap between the current rule-based workflow and the agent-ready version is your roadmap. It tells you exactly what’s missing: data quality, governance, decision authority, or all three.

The architectural shift Scott envisions will probably arrive. The question isn’t whether the direction is right. It’s whether marketing’s strategic interests get designed into the foundation or bolted on after someone else builds it. The vocabulary lesson starts now. Not because the architecture is ready, but because the decisions that shape it are already being made.