Marketing Has an AI Problem, and It Has Nothing to Do with AI

Marketers have nearly doubled their AI use in the last two years, but performance is still not being optimized.

That’s the uncomfortable finding in the 35th edition of The CMO Survey, based on responses from 308 senior marketing leaders. AI adoption is real and accelerating. Companies now expect it to power more than half of all marketing activity within three years. Content creation (73.9%), personalization (65.4%), automation (48.9%), data analysis (46.3%), and targeting (45.2%) have all seen strong adoption growth since 2023. Even generative engine optimization (GEO), getting content surfaced in AI search results, is already used by 41.5% of companies, remarkable for a tool that barely existed two years ago.

The business case looks good on paper too. Companies using AI report gains in sales productivity (+14.1%), customer satisfaction (+10.8%), and marketing overhead reduction (−14.6%), and those numbers have been climbing year over year.

Against this positive backdrop, there are troubling indicators that technology adoption has outpaced the organizational readiness needed to maximize its impact. When asked to rate how well your company is performing each of the following marketing technology activities, including things like “integrating marketing technologies into our funnel” and “generating ROI from marketing technologies,” no activity breaks 5 on a 7-point performance scale (1 = “poorly,” 7 = “very well”). Worse, no rating has improved in two years, despite all the investments companies have made in AI. When we asked marketers what’s holding AI back, three of the top four answers shown below had nothing to do with the technology: lack of budget, talent management, and managerial bandwidth.

The question isn’t whether AI works in marketing anymore. It’s whether marketing organizations are building the conditions that let it work.

Where Leaders Go Wrong

Five forces are working against AI’s impact in most marketing organizations, and they feed each other.

You’re shrinking the team you need to grow. Training and development budgets have fallen to 3.8% of marketing budgets, down from 5.8% in 2019. Headcount growth has slowed by more than half compared with last year, with the largest companies actively shrinking their marketing teams. Leaders rate their own “hiring to manage marketing technologies” and “training employees on emerging technologies” among their weakest activities, at 3.7 and 3.9 on a 7-point scale. One CMO Survey respondent put it bluntly: “We too often have leaders that ask good questions, but don’t have enough base knowledge of technology to lay out a qualified strategy.” AI doesn’t run itself. Cutting the people who use it while starving development of the ones who remain is a losing trade.

Your capability strategy hasn’t caught up to what AI demands. Nearly 60% of marketing leaders still build AI capability internally through hiring and training rather than partnering or acquiring it, a number that hasn’t budged since 2020, despite six years of the ground shifting under marketing’s feet. Given the pace of change illustrated in the charts below and the specialized expertise AI requires, a “build everything ourselves” strategy is increasingly the wrong bet. A more aggressive partnership approach, or at least a more balanced mix, is likely to close capability gaps faster than internal development alone.

AI is living in silos, not systems. Companies pour investment into AI for content creation (74% adoption) while leaving other parts of the funnel untouched. CPG companies, for instance, use AI for content creation 64.7% of the time but for targeting only 29.4%. Compare that to real estate, which splits AI use more evenly (60% content, 80% targeting) and reports a 30% jump in sales productivity, the best of any sector. One respondent named the core issue directly: “The biggest barrier is the fragmentation of systems … having too many tools that create complexity and don’t always integrate or communicate effectively with each other.” AI pays off when it’s deployed as a connected system, not a pile of point solutions.

Your C-suite partnerships are too weak to fund any of this. The CMO-CFO relationship scores just 4.5 out of 7 for building a credible business case for marketing spend, barely up from 4.3 in 2021. Only 55.7% of companies report growing collaboration between marketing and IT leadership, even as AI’s technical demands increase. Without a strong CFO relationship, capability investment loses out to budget pressure every time. Without a strong CTO relationship, marketing can’t get the technical architecture AI needs to work.

You’re using AI to cut costs when you should be using it to grow. Marketing leaders report spending two-thirds of their time managing the present and only one-third preparing for the future, a ratio that hasn’t shifted since 2019. That short-term orientation shows up directly in the gains AI is producing: Cost-reduction gains are outpacing the gains in sales productivity and customer satisfaction, as shown in the charts below. One respondent called it a “vending machine view by owners,” a constant demand for AI to prove immediate ROI or get cut. It shows: Only 56.5% of companies say they’re effectively using technology to pursue growth, down from 2023.

The Compounding Problem

None of these five forces operate alone. Short-termism kills appetite for AI investment. Reduced investment pushes AI into silos instead of systems, which limits its value and weakens the case for investing further, fueling more short-termism. Weak C-suite partnerships keep the cycle intact. The “build” strategy persists even as the resources to execute it erode. And narrowing ambition means that even the AI that does get deployed well is underused.

What Leaders Should Do

Protect the human layer first. Capability investment isn’t a discretionary cost to cut when profits disappoint; it’s part of the AI investment itself. The data backs this up starkly: Education spends just 1.2% of its budget on training and posts some of the weakest AI returns of any sector. Real estate invests the most in training and saw a 30% jump in sales productivity. Make the conversation with your CFO about the cost of skipping capability investment, not the cost of making it.

Stop assuming you have to build it yourself. For AI-specific gaps, analytics, generative AI, and GEO, partnering makes more sense than building. The specialized knowledge required and the speed at which it’s changing make pure internal development a losing strategy, especially as the resources to pull it off keep shrinking.

Integrate, don’t just adopt. The gap between content creation and targeting isn’t a technology problem; it’s a strategy problem. Closing it means planning deliberately for how AI-generated content reaches the right audiences, and building the CTO partnership needed to connect data across every touchpoint.

Sell AI as a growth investment, not cost-cutting. The data is clear that AI drives real gains in sales productivity and customer satisfaction. Leaders who pitch it to their CFOs primarily as a cost-reduction tool are making their own budgets easier to cut.

Expand your ambition. AI can’t reach its full potential in an organization that has narrowed its focus to existing customers and existing markets. Getting more out of AI starts with asking more of your organization, and of yourself.

The Bottom Line

The evidence is no longer speculative: AI is delivering real, measurable value in marketing, and that value is growing year over year. But the technology was never the hard part. The hard part is the organizational, financial, and strategic choices being made around it, and right now, most marketing organizations aren’t making the ones that let AI live up to what it’s already capable of.

Based on findings from the 35th edition of The CMO Survey.