Investment Is Rising, But the Hard Decisions Haven't Been Made Yet
From Experimentation to Commercial Advantage
There is no shortage of CMO ambition around AI. The budgets are moving, the conversations are happening, and the optimism is genuinely high.
But ambition and operating model are two very different things. Right now, most marketing functions have a lot more of the former than the latter.
Understanding where CMOs are actually putting their weight helps CEOs ask better questions — not "are we investing in AI?" but "are we investing in the right things, for the right reasons, in the right sequence?"
How Much Are CMOs Actually Spending on AI — and Is It Working?
BCG's 2025 CMO research found that
- 71% of CMOs plan to invest more than $10 million annually in GenAI over the next three years, up from 57% the year before.
- Optimism among CMOs has risen from 74% to 83% in the same period.
That is not a tentative commitment.
For larger organisations, AI has moved from a test-and-learn conversation to a strategic investment decision — treated the same way a major CRM implementation would be.
For mid-market and growing businesses, the investment threshold is lower but the logic is the same. Deloitte found that two thirds of companies had already adopted three or more AI use cases, concentrated in content creation, predictive analytics and conversational AI.
The important caveat: Gartner found
only 30% of CMOs have the data foundations, governance and team capability to use that investment properly.
The investment is coming. The infrastructure, in many cases, is not yet there.
What Are CMOs Prioritising AI For — and What Does the Data Say Is Actually Working?
The honest picture of where AI sits in most marketing functions today is this: heavily weighted toward content and creative production, moving into reporting and analytics, beginning to touch personalisation, and still largely aspirational when it comes to customer journey orchestration and agentic workflows.
Content and creative is where almost every team started, because the friction was obvious and the application was immediate. Gartner found that
among CMOs already using GenAI, 77% are using it for creative development tasks.
The productivity case is real. Mondelez reported a 30% to 50% reduction in content production costs through a purpose-built AI content tool — with human review and brand safety rules built in from the start. That is a meaningful saving, but it took deliberate process design, not just a tool subscription.
The more significant shift is happening in analytics and reporting. Nearly half of CMOs using AI report large benefits in campaign evaluation and reporting. Better measurement changes what gets decided, not just how fast reports get produced. And that is where marketing's commercial influence either grows or stalls.
How Is the CMO Role Changing Because of AI?
The CMO role is shifting — and AI is accelerating a change that was already underway.
For years, the CMO role drifted toward campaign leadership and performance accountability: managing channels, agencies and spend. AI is not taking that work away. But it is forcing a more important question about what CMOs should actually be spending their time on.
If AI handles a significant portion of content production, reporting summarisation and basic segmentation,
the CMO's real value is not in those tasks. It is in the decisions that sit above them.
Which customers matter most, and why. What the proposition needs to say, and to whom. Where demand is actually coming from and how to build it sustainably. How marketing connects to the commercial growth model of the business.
BCG's research framing is useful: the CMO role is becoming less about campaign leadership and more about operating model leadership.
- What workflows change.
- What gets automated.
- What stays human-led.
- What the governance looks like.
- What gets measured and how.
That is a more demanding role, not an easier one. And it requires a different kind of involvement from the CEO.
What Should a CEO Be Asking Their CMO About AI — That Most CEOs Are Not Asking?
If your marketing leader is presenting an AI strategy that is primarily about tools — which platforms to use, which content tasks to automate, which agency to work with — it is worth asking a harder question.
Which marketing decisions will be better because of AI — not faster, better?
The distinction matters. AI that speeds up poor decision-making is not an asset.
What does the data picture look like?
Before any significant AI investment in personalisation or customer journey, the CEO needs to understand whether the underlying customer data is unified, reliable and accessible. If not, the AI roadmap should start there.
How is marketing measuring commercial outcome — not just activity?
Productivity improvements are a starting point, not the commercial measure. The right question is whether AI is improving customer understanding, conversion, retention or revenue.
Where is human judgement non-negotiable?
AI can produce content, generate campaign variants, summarise performance data and suggest next steps. It cannot replace the strategic and commercial judgement that ensures those outputs are actually right for the customer, the brand and the business.
What Is the Right Way to Think About AI as a Marketing Operating Model — Not Just a Tool?
The businesses furthest ahead in two years will not be the ones that invested most heavily in AI tools. They will be the ones that redesigned their marketing operating model around AI capability early enough to get it right.
That means deciding deliberately where AI sits in the workflow and what it is — and is not — responsible for. It means training the team not just to use tools, but to review, validate and improve AI output with commercial judgement. It means building governance that is light enough to move quickly and structured enough to manage risk.
CMOs thinking this way are not asking "which AI tools should we buy?" They are asking
What does a well-governed, AI-enhanced marketing function look like for a business at our stage, with our data, serving our customers?
That is the more useful question. And it starts with leadership clarity, not a software subscription.
Why are CMOs investing so much in AI when only 30% have the readiness to use it?
Competitive pressure is moving faster than organisational readiness. The risk is spending on capability before building the foundations that make it work — data quality, governance and team skills. The businesses getting the best return are investing in the foundations first.
What is the difference between AI that helps CMOs and AI that transforms marketing?
AI that helps CMOs handles tasks: drafting content, summarising reports, generating variants. AI that transforms marketing changes decisions: which customers to prioritise, how to allocate budget, where demand is coming from, what the proposition needs to say. Most companies are in the first category. The second is where commercial advantage is built.
How should a CEO evaluate whether their CMO's AI strategy is sound?
Ask three questions: Does it connect to a specific commercial problem? Does it account for the quality of the data underpinning it? Does it include a clear governance model for who reviews and approves AI output? If any of those three are missing, the strategy is incomplete.
Catherine Mak is Founder of ZILU consultancy. She works as a #FractionalCMO with growth-stage businesses, scaleups and organisations navigating AI adoption in their marketing functions. Clients include Citibank, Bupa Global and PEI Group.