Commercial Operations Orchestration (CO2)

Table of Contents

  1. Definition
  2. Why it matters for marketing
  3. How it works
    1. How CO2 is measured
  4. How to utilize CO2
  5. Comparison: CO2 vs. related operating models
  6. Best practices
  7. Future trends
  8. FAQs
  9. Related Terms
  10. Listen: related episodes of The Agile Brand
  11. Sources

Definition

Commercial Operations Orchestration (CO2) is an operating model that runs marketing, sales, service, and product as one commercial system instead of a chain of departments passing customers to each other. Humans and AI agents act from a shared view of the customer’s current state, follow one set of commercial policies, and get measured on outcomes across the whole customer lifecycle rather than on departmental throughput.

Greg Kihlström introduced the term in Marketing Operations 3.0 (De Gruyter Brill). There, CO2 is described as the structural response to a simple mismatch: customers experience one company, but most companies still operate as a collection of departments with their own systems, definitions, and clocks.

See also: Revenue Operations (RevOps) · Marketing Operations · Customer Journey Orchestration (CJO) · AI Agent

Not to be confused with: “Commercial operations” as a job function (common in pharma and life sciences, where it covers sales operations, analytics, and market access), or CO₂ as carbon dioxide in sustainability reporting.

Back to Top

Why it matters for marketing

For most of its history, marketing operations sat inside marketing. It launched campaigns, routed leads, connected platforms, and produced reports. RevOps widened that scope to sales and customer success, but it stayed mostly a front-office alignment model built around the pipeline.

AI agents put pressure on that arrangement. Once agents take part in outreach, qualification, negotiation, service, and retention, every handoff between teams becomes a place where context gets lost. A customer doesn’t care whether a message came from a marketing workflow, a sales queue, or a service platform. If those systems disagree, the brand looks inconsistent, and an agent working from a stale record will act on the wrong facts at machine speed.

CO2 moves the operations leader’s job from managing campaigns and platforms to managing commercial continuity. The question shifts from “Is our automation working?” to “Can the company sense, decide, and respond across the customer lifecycle fast enough to compete?” That’s a different job description for marketing ops, and it puts the function closer to the center of how the business grows.

Back to Top

How it works

The book describes CO2 as resting on four requirements:

  1. A shared truth layer. One current record of the customer’s commercial state (recent friction, commitments already made, consent, product usage) that every team and agent reads from, instead of each function rebuilding its own version.
  2. Cross-functional decision logic. Rules and models that operate across commercial moments, not inside one department’s queue. A billing correction, for example, can inform a renewal offer without waiting for a handoff.
  3. Policy clarity. Machine-readable limits on what agents may do, such as discount ceilings, escalation triggers, and restricted actions.
  4. Governance with real authority. A cross-functional body (the book calls it the CO2 Council) that owns decision rights over policy, ownership, investment, and acceptable autonomy.

Work moves through a common orchestration layer rather than from team to team. That’s what lets the system respond in parallel. Marketing, sales, and service act on the same context at the same time instead of each waiting to inherit a simplified version of it.

The basic working unit is the human-agent squad. One human lead sets goals, guardrails, and escalation conditions, and a small group of specialized agents handles execution. The book’s “1-to-5 model” pairs one human with roughly five agents.

Back to Top

How CO2 is measured

CO2 doesn’t have its own formula. The book measures it with Return on Agent (ROA), applied as a shared KPI across functions:

ROA = (Value Created by Agented Work − Total Cost of Agented Work) / Total Cost of Agented Work

Making ROA a shared score is the point. No single team can claim success while the full customer experience degrades, and friction can’t be pushed into another team’s queue and called efficiency.

Back to Top

How to utilize CO2

Common starting points:

  • Map signal loss first. Trace one high-value journey (trial to paid, say, or onboarding to first renewal) and record every place context is dropped, re-entered, or reinterpreted between teams.
  • Pick one cross-functional outcome. Early churn prevention or onboarding continuity work well, because both depend on marketing, service, and product data at once.
  • Stand up one human-agent squad around that outcome, with a named human lead and clearly bounded agent authority.
  • Write policies down in machine-readable form before expanding agent autonomy. A discount limit that only lives in a sales deck can’t constrain an agent.
  • Bring product into the loop. Usage, adoption, and support signals should change roadmap and offer decisions, not just show up in a quarterly review.
  • Form the governing council once more than one squad is running, so decisions about shared policy don’t fall back to whichever department shouts loudest.
Back to Top
Marketing Operations (2.0)Revenue Operations (RevOps)Customer Journey Orchestration (CJO)Commercial Operations Orchestration (CO2)
ScopeMarketing functionMarketing, sales, customer successCustomer-facing interactions across channelsMarketing, sales, service, product, and post-sale value
Primary unitCampaignPipeline / funnel stageJourney or interactionCustomer’s shared commercial state
How work movesRules-based workflowsAligned handoffs between teamsReal-time decisioning per interactionParallel action from shared context
Role of AI agentsTool-like automationAssistive (forecasting, scoring)Decisioning enginesAutonomous execution within defined policy
Core metricCampaign ROI, MQLsPipeline, bookings, forecast accuracyEngagement, conversion per journeyReturn on Agent (shared across functions)
GovernanceMarketing leadershipRevOps leader, often under the CROUsually marketing or CXCross-functional council with decision rights

CJO is a capability or platform category. CO2 is an operating model, and a CJO platform could be one piece of its orchestration layer.

Back to Top

Best practices

  • Treat CO2 as an operating model change. Buying a new integration platform won’t produce it.
  • Keep functional expertise intact. Teams keep their specialties and reporting lines; what changes is that they act from shared context and are accountable for shared outcomes.
  • Define “value,” “churn risk,” and “commercial state” once, and use those definitions everywhere. Inconsistent definitions are one of the quickest ways a shared system splits back into silos.
  • Include service as part of the commercial loop. A support interaction is also a renewal signal.
  • Set escalation thresholds before granting autonomy. Decide in advance which moments (pricing exceptions, renewal disputes, reputationally sensitive replies) go to a human.
  • Separate experimental systems from production systems so pilots don’t quietly become ungoverned operations.
  • Watch for superficial progress. A good demo of an agent isn’t the same as a governed capability with clear authority and drift detection.
Back to Top
  • More agent-to-agent interaction. As customers delegate comparison and negotiation to their own AI agents (see B2A2C), the brand’s internal coherence becomes visible to outside machines, and inconsistencies get caught faster.
  • Governance encoded in systems. Policies, authority limits, and prohibited actions are moving into machine-readable rules that are checked continuously rather than reviewed after the fact.
  • New roles. Expect titles such as Bot Manager, Agent Orchestrator, or CO2 Lead, focused on supervising agent performance rather than executing tasks directly.
  • Shared ROA-style metrics replacing departmental scorecards in executive reporting.
  • Rising cost of ungoverned AI. Forrester’s 2026 B2B predictions forecast that B2B companies will lose more than $10 billion because of ungoverned use of generative AI. That’s a strong argument for the policy and governance layers CO2 requires.
Back to Top

FAQs

What does CO2 stand for? Commercial Operations Orchestration. The acronym comes from Greg Kihlström’s book Marketing Operations 3.0.

Is CO2 just a new name for RevOps? No. RevOps aligns marketing, sales, and customer success around shared pipeline data. CO2 extends to service, product, and post-sale value, and it’s built for a workforce in which AI agents execute work, not just assist with it. The book describes RevOps as an important step but not a sufficient destination.

Who owns CO2 in an organization? Ownership sits with a cross-functional governing body (the CO2 Council) rather than one department. Marketing operations is often where it starts, because that function already manages much of the data and automation involved.

Does CO2 require reorganizing departments? Not necessarily. Functions can keep their structure and expertise. What changes is the operating logic: shared context, shared policy, and shared accountability for outcomes.

What technology does CO2 need? A data layer that keeps a current, consistent customer state across systems, plus machine-readable policies agents can act within. A Customer Data Platform or data warehouse often plays part of that role. No single product delivers CO2 on its own.

How is CO2 measured? With Return on Agent (ROA), used as a shared KPI across functions, backed by driver metrics (such as escalation quality and value realization) and guardrail metrics (such as compliance exceptions and drift rate).

Is CO2 only for B2B companies? No. The book’s running case study is a consumer telecom and home-internet provider. Any business where marketing, sales, service, and product all touch the same customer can apply it.

Where should a company start? Pick one journey where context gets lost between teams, measure the delay and rework it causes, then build a single human-agent squad around fixing it before scaling.

How does CO2 relate to customer experience? Directly. Most of the friction customers feel (repeating themselves, conflicting answers, waiting for systems to sync) comes from handoffs between functions. CO2 is designed to remove those handoffs.

Back to Top
  1. Revenue Operations (RevOps)
  2. Marketing Operations
  3. Customer Journey Orchestration (CJO)
  4. Next Best Action
  5. Business-to-Agent-to-Consumer (B2A2C)
  6. Human-in-the-Loop (HITL)
  7. Decision Rights
  8. Structural Friction Index (SFI)
  9. Return on Agent (ROA)
  10. CO2 Council
Back to Top
Back to Top

Sources

Back to Top

Tags:

Was this helpful?

Previous

Business Value

Next

Contract Lifecycle Management (CLM)