The ACO Maturity Model

Part 1 argued that the brands winning in agentic commerce treat ACO as a function rather than a project, and laid out the five operating layers a mature function runs on: strategy, ownership, workflow, governance and measurement, and systems.

That is a frame. It does not tell you where your organization actually sits today. Most commerce leaders we talk to already know they have gaps in one or two of those layers. What they want is a clear way to see where they are, what good looks like at the next stage, and what to focus on to get there. That is what this piece is built to provide.

Three stages, defined by alignment

The ACO Maturity Model has three stages: Explorer, Builder, and Leader. Each stage is defined by how aligned the five operating layers are, not by how many people work on ACO. Two organizations with the same headcount can sit in different stages. A two-person team operating with a clear owner, a defined workflow, policy-level governance, and a system layer behind them is functionally further along than a five-person team running ad-hoc on manual processes.

The system layer is the dividing line

Of the five operating layers, the one that most predictably separates stages is systems. Explorer-stage organizations typically have no system layer at all, which is why their ACO work cannot scale beyond a curated subset of the catalog regardless of how skilled or motivated the people doing it are. Builder-stage organizations have it partially in place, supporting some of the work but still leaving meaningful gaps. Leader-stage organizations run continuous execution and measurement across the full catalog as a baseline.

This is not a tooling pitch. It is a structural observation. ACO across every AI-driven channel, at catalog scale, requires continuous execution, context capture, and optimization across thousands of SKUs simultaneously. That is operationally impossible without a system layer doing the continuous work. Strategy, ownership, workflow, and governance can be fully aligned, and the function will still cap out without it.

Where do you stand today?

Four questions, each tied to one of the operating layers most diagnostic of maturity: ownership, workflow, governance, and systems. There is no strategy question in the assessment, not because strategy does not matter, but because strategy is the easiest layer for organizations to self-report inaccurately. The other four reveal what is actually true about how the function operates.

What each stage actually looks like

Once you have a stage, the question becomes what good looks like there, what good looks like at the next stage, and what is realistic in the next two quarters. The detailed descriptions below are drawn from how the commerce organizations we work with operate at each stage today.

Explorer · Ad-hoc

ACO sits inside an existing SEO, digital commerce, or catalog role. Typically one person, part-time, working through a limited set of products. Workflows are manual. Governance is informal. There is usually no system layer, which means the work cannot scale beyond a small subset of the catalog. This is where most organizations are today.

McKinsey's finding that only 1% of companies have reached AI maturity, paired with BCG's 5% "future-built" benchmark, suggests that the long tail of Explorer-stage organizations represents a real competitive opening. The most common failure modes at this stage are the same failure modes seen in early agentic AI deployments more broadly: introducing AI into environments with significant technical debt, allowing siloed teams to spin up duplicative agents without coordination, and using AI to automate existing processes rather than rethink them. Each is a sign the organization is treating ACO as a project rather than a function.

Builder · Structured

ACO ownership is emerging. Two to four people contribute as part of broader roles, drawn from catalog and data, visibility and discovery, and content and brand disciplines. Workflows are defined and starting to be systematized. Initial governance guidelines and approval frameworks exist but are still evolving. The system layer is partially in place.

This is where most organizations need to land within the next year if they want to compete in the 2026 holiday cycle and beyond. The Builder stage is achievable in months when an organization commits to it, and the move from Explorer to Builder is the one that unlocks the most value because it is where the function shifts from individual effort to coordinated capability.

Leader · Scaled

ACO operates as a dedicated function. Some organizations carve out their own P&L for it, mirroring how retail media budgets were eventually split out from broader digital marketing. The team spans six disciplines, strategy and leadership, product data and catalog, visibility and discovery, content and brand compliance, technical integration, and measurement and analytics, all coordinated under a single-threaded owner.

Workflows are continuous and system-driven, with human-in-the-loop review focused on exceptions rather than every change. A formal governance model, often called an AI Council or equivalent, makes policy-level decisions that the system enforces consistently at SKU level. Coverage and ownership are measured continuously across every agentic commerce channel.

The move from Explorer to Builder is the one that matters

The maturity model is not a sequence everyone has to walk through manually. Leader stage is the destination, but the move from Explorer to Builder is the one that unlocks the most value and is the most achievable in the near term. It is also the move that determines whether an organization is competing in the 2026 holiday cycle or watching it from the outside.

The first step is always the same: name the single-threaded owner. That move alone changes how the work coordinates, how governance forms, and how the system layer eventually gets selected and rolled out. Everything else in Part 3 builds on that decision.