Frameworks

The following frameworks were developed in The Agentic CMO to give marketing leaders practical models for navigating the transition from automation to autonomy. Each addresses a distinct strategic challenge — from assessing AI maturity to designing hybrid organisations to scaling transformation across the enterprise.

Framework 01

The PACE Model

Perception, Action, Reasoning, Learning

The PACE Model provides a structured lens for evaluating agentic AI systems and an organisation's readiness to deploy them. Rather than treating AI capability as a single dimension, PACE decomposes it into four interdependent facets. Perception measures how effectively a system senses and ingests signals from its environment — customer behaviour data, market shifts, competitive signals, internal performance metrics. An AI system with weak perception operates on stale or incomplete information, regardless of how sophisticated its other capabilities may be.

Action assesses the system's capacity to execute decisions autonomously — not merely to recommend, but to act. This is the critical differentiator between traditional analytics and agentic AI. Reasoning evaluates how the system processes information, weighs trade-offs, and arrives at judgements. Advanced reasoning enables an agent to handle ambiguity, reconcile conflicting objectives, and operate within strategic guardrails without constant human oversight. Learning captures the system's ability to improve through feedback loops — adapting its behaviour based on outcomes rather than requiring manual recalibration.

Together, the four dimensions create a maturity assessment that helps organisations identify where they sit on the autonomy spectrum and where investment will yield the greatest return. A system strong in perception and reasoning but weak in action, for example, is an expensive advisory tool rather than a true agent. The model enables CMOs to move beyond binary "we have AI / we don't" conversations toward nuanced capability planning.

Explore the PACE Model in The Agentic CMO →

Framework 02

The Agentic CMO Competency Model

12 competencies across 4 domains

The transition to agentic marketing demands new leadership capabilities that most CMO development programmes do not address. The Agentic CMO Competency Model defines twelve specific competencies organised across four domains: AI Literacy, Change Leadership, Ethical Governance, and Technical Fluency. The model serves as both a self-assessment tool for individual leaders and a framework for building leadership development programmes at scale.

AI Literacy encompasses the ability to evaluate AI capabilities critically, understand the difference between deterministic and probabilistic systems, and communicate AI strategy to boards and stakeholders without resorting to jargon or hype. Change Leadership covers the human dimension — managing resistance, building coalition support, redesigning roles without destroying morale, and maintaining momentum through the inevitable setbacks of transformation. Ethical Governance addresses the CMO's expanding responsibility for algorithmic bias, data privacy, transparency, and the societal impact of autonomous marketing systems. Technical Fluency — distinct from technical expertise — ensures leaders can engage meaningfully with engineering teams, evaluate vendor claims, and make informed architectural decisions.

The model matters because the gap between AI-capable organisations and AI-leading organisations is almost always a leadership gap, not a technology gap. Organisations with the right technology but the wrong leadership competencies consistently underperform those with adequate technology and exceptional leadership. The competency model gives organisations a concrete framework for closing that gap.

Explore the Competency Model in The Agentic CMO →

Framework 03

The Value Stack

Efficiency → Enhancement → Strategic Value

Most organisations begin their AI journey by pursuing efficiency gains — automating repetitive tasks, reducing manual effort, cutting production costs. This is the base layer of the Value Stack, and it delivers real, measurable returns. Mondelez achieved 30–50% reductions in creative production costs; Walmart improved operational efficiency by 25%. But efficiency alone is a cost play, not a strategic one.

The second tier, Enhancement, represents AI that improves the quality of decisions and outputs rather than merely accelerating existing processes. Starbucks' Deep Brew platform operates at this level — not just sending more emails faster, but making millions of autonomous personalisation decisions daily across 140,000+ weekly permutations, each one better than a human marketer could achieve at that scale. Enhancement creates competitive differentiation because it changes the nature of the output, not just the speed.

The third tier, Strategic Value, is where agentic AI generates entirely new sources of competitive advantage — new business models, new revenue streams, new forms of customer relationship that were previously impossible. Coca-Cola's Create Real Magic platform, with 150,000+ AI-generated creative variations, represents early movement into this tier. The Value Stack helps leaders communicate to boards and CFOs that AI investment is not a single line item with a single return profile, but a progressive capability build where the most transformative value comes from sustained investment through all three tiers.

Explore the Value Stack in The Agentic CMO →

Framework 04

Three Architectural Models

Hub-and-Spoke · Embedded · Hybrid Collaborative

How an organisation structures the relationship between human teams and AI systems is as consequential as which AI systems it deploys. The Agentic CMO identifies three primary architectural models. The Hub-and-Spoke model centralises AI capabilities in a dedicated team (the hub) that serves multiple business units and functions (the spokes). This model excels at maintaining consistency, building deep technical expertise, and ensuring governance standards are uniformly applied. It works best for organisations early in their AI journey or those operating in heavily regulated industries where centralised control is essential.

The Embedded model distributes AI capabilities directly within each functional team. Marketing, sales, customer service, and product teams each develop or manage their own AI agents. This model maximises speed and domain relevance — teams closest to the customer build the AI that serves the customer. However, it risks fragmentation, inconsistent governance, and duplicated effort. It suits organisations with strong existing team autonomy and mature data infrastructure.

The Hybrid Collaborative model combines elements of both — a central team owns the platform, standards, and governance layer, whilst functional teams build and customise agents within those guardrails. This is the model the book ultimately advocates for most large enterprises, as it balances innovation velocity with enterprise-grade control. The choice between models is not permanent; organisations typically evolve from hub-and-spoke toward hybrid collaborative as their maturity increases.

Explore the Architectural Models in The Agentic CMO →

Framework 05

The 90-Day Agentic Transformation Plan

Three phases from assessment to scale

The final part of The Agentic CMO translates strategy into action with a structured 90-day plan. Phase 1 (Days 1–30): Assessment and Foundation focuses on auditing the current state — mapping existing AI capabilities against the PACE Model, assessing leadership readiness against the Competency Model, identifying quick wins, and establishing the governance framework that will guide all subsequent deployment. This phase ends with a transformation roadmap and executive alignment.

Phase 2 (Days 31–60): Pilot and Prove moves from planning to execution with targeted pilots designed to generate measurable results quickly. The emphasis is on selecting pilot use cases that are strategically significant — not trivial demonstrations that fail to build organisational conviction. Each pilot is structured with clear success metrics, risk boundaries, and feedback loops that feed into the learning cycle. This phase generates the evidence base needed to secure broader investment.

Phase 3 (Days 61–90): Scale and Embed transitions successful pilots into production systems, establishes ongoing performance measurement against the Value Stack, and builds the organisational muscle memory for continuous AI integration. The plan is deliberately designed to move fast enough to maintain executive momentum whilst building sufficient governance to avoid the common pitfalls of premature scaling. It draws heavily on BCG's S-curve model to set realistic expectations about the timeline from pilot to enterprise-wide value creation.

Get the full 90-Day Plan in The Agentic CMO →

Framework 06

BCG S-Curve Model Application

From pilot purgatory to enterprise scale

One of the most persistent problems in enterprise AI is what the book calls "pilot purgatory" — organisations that run endless proofs of concept without ever achieving production-scale deployment. The Agentic CMO applies BCG's S-curve model to explain why this happens and how to break through. The S-curve describes the characteristic pattern of value creation from new technology: a slow initial period of investment and learning (the flat bottom of the S), followed by rapid acceleration as the organisation builds capability and confidence (the steep middle), before eventually plateauing as the technology matures (the flat top).

Most organisations stall at the bottom of the S-curve because they underinvest in the foundational capabilities — data infrastructure, governance frameworks, talent development, organisational redesign — that enable the steep acceleration phase. They chase quick wins at the efficiency layer of the Value Stack without building the platform for enhancement and strategic value. The book argues that the transition from the flat bottom to the steep middle requires a deliberate investment in organisational capability that goes far beyond technology procurement.

The practical application involves mapping each AI initiative against the S-curve to set realistic value expectations, identifying the specific capability gaps that are preventing acceleration, and designing investment portfolios that balance short-term efficiency gains (to fund the journey) with medium-term capability building (to reach the steep section). This framework is particularly valuable for CFO conversations, where unrealistic expectations about AI ROI timelines are a primary cause of programme cancellation.

Explore the S-Curve application in The Agentic CMO →