80% of Marketers Are Under Pressure to Adopt AI. Only 6% Have Actually Done It. Here's Why.
/The number that should stop every marketing leader in their tracks: 80% of marketers are under pressure to adopt AI, but only 6% have fully embedded it into their workflows.
That gap — between the pressure and the reality — is what the Supermetrics 2026 Marketing Data Report found when it surveyed 435 marketers across brands and agencies globally. I'm quoted in that report. I also live this tension every day as CMO at Supermetrics, a platform that sits at the centre of how marketing teams connect, manage, and activate their data.
AI adoption in marketing is failing almost everywhere, and the reason isn't the technology. It's that most marketing functions are trying to adopt AI on top of foundations that can't support it — fragmented data, absent strategy, and a structural disconnect between where marketing decisions are made and where data strategy is set. The 6% who have fully embedded AI into their workflows fixed those foundations first. The 94% haven't.
This piece is about what separates the 6% from the 94%, and what marketing leaders need to fix before any AI investment makes sense.
The pressure is real. The strategy isn't.
The 2026 Marketing Data Report makes one structural problem very clear: AI adoption is being driven from the top, but the strategy to support it isn't following.
89% of marketers say pressure to adopt AI is coming from the C-suite and board. That's not unusual — boards have been asking about AI for two years. What's unusual is that 37% of those same marketers say they lack a clear AI strategy from leadership. The pressure is coming from the same place the strategy should come from, and it isn't landing.
The result is what happens when organisations conflate adoption with implementation. Teams buy tools. Individuals experiment. Pockets of activity emerge. But without a defined use case, a connected data model, and a measurement framework to evaluate what's working, those experiments stay experiments. They don't compound. They don't scale. And they don't show up in the 6%.
There's also a trust problem: 39% of marketers cite concerns about AI data privacy as a barrier. That's a legitimate concern, but it's also a symptom of the same underlying issue. When data governance is unclear and data ownership is fragmented, every AI use case becomes a risk question rather than an opportunity question.
AI is being deployed where it's easiest, not where it matters most
The report is specific about where AI is actually being used: 87% of marketers are using it for content creation, copywriting, and creative ideation. Only 39% are using it for reporting and analytics. Only 33% for marketing automation.
Content creation is a reasonable starting point. It's low-risk, accessible, and produces visible output quickly. But it's also the lowest-leverage application of AI in a marketing function. It accelerates the production of something that was already happening. It doesn't change what the function can know, decide, or do at a structural level.
The applications that change what a marketing function can actually do — connecting data across channels, automating performance analysis, enabling real-time optimisation, generating insight rather than content — are the ones sitting at 33–39% adoption. Those are the use cases where AI compounds. Where the 6% live.
As Marianna Imprialou, Head of Data Science at Supermetrics, noted in the report: "The risk is that AI becomes pigeonholed as nothing more than an accelerator of basic tasks. That feels like a very real outcome unless organisations start thinking more creatively and embedding AI into complex, higher-value work." That outcome is already visible in most marketing functions today.
The data problem nobody wants to name
Here is what I believe is the single most important finding in the report, and the one that gets the least attention in conversations about AI adoption.
52% of marketers say data strategy and measurement decisions are made by external teams — IT, data engineering, or some combination of both. Only 31% of CMOs are meaningfully involved in data strategy decisions.
That is not a data problem. It is a marketing leadership problem.
When marketing doesn't own or meaningfully shape its data strategy, several things happen in sequence. The data that gets prioritised is the data that's easiest to collect, not the data that's most useful for marketing decisions. Feedback loops between campaign performance and campaign decisions slow down because the people who need the data aren't the people who manage it. And AI adoption stalls because every use case requires a data conversation that marketing doesn't control.
The 2026 report frames it directly: a more useful way to think about marketing data is not as a source of certainty, but as an input to judgment. Data is evidence, not truth. Metrics are signals, not decisions. That distinction only becomes actionable when marketing leaders are close enough to the data to apply it — and 52% of marketing functions have ceded that ground.
When data ownership is blurred, marketing either overreaches into infrastructure or gets pushed downstream into communications execution. Neither position supports AI adoption at scale.
What the 6% are doing differently
The report doesn't profile the 6% in isolation, but the data points to what separates them from the majority. Based on the findings and what I've seen in practice, four things stand out.
They started with the data foundation, not the tools
The marketers who have fully embedded AI into their workflows didn't start by buying AI tools. They started by getting their data into a state where AI could do something useful with it — clean, connected, governed, and accessible. The report is unambiguous on this: you can't outrun a bad data foundation. Clean, connected, and governed data is the prerequisite for AI adoption, marketing measurement, and personalisation at scale. The 6% understood that before they started.
They defined use cases before deploying tools
The majority of AI adoption is happening in silos, driven by individual enthusiasm rather than defined business problems. The 6% worked backwards — from the decision they needed to make or the outcome they needed to improve, to the data required to support it, to the AI capability that could close the gap. That sequence matters. AI deployed in search of a use case rarely finds one that scales.
Marketing is meaningfully involved in data strategy
Among the organisations where AI adoption has progressed furthest, marketing isn't waiting for data to be delivered by another team. The CMO or a senior marketing leader is in the room where data strategy is set — shaping the priorities, the definitions, and the governance framework. That involvement isn't just politically important. It's practically necessary. The use cases that matter most for marketing are the ones that marketing leaders are best placed to define.
They treat ROI as a signal, not a calculation
40% of marketers in the report struggle to prove ROI across channels. Among the 6%, ROI isn't treated as a precise calculation that validates AI investment after the fact. It's treated as a directional signal — a collection of indicators that, taken together, tell you whether the function is moving in the right direction. That's a more honest and more useful relationship with measurement, and it's the one that allows AI investment to be defended at board level without overpromising on precision.
What CMOs need to do before the next AI conversation
Most CMO-level conversations about AI are happening at the wrong level. They're about tools, vendors, and use cases. They should be about foundations, ownership, and strategy.
Before the next AI investment decision, three questions are worth asking with honesty.
Does marketing own or meaningfully shape its data strategy? If 52% of functions have ceded this ground, the first move isn't an AI tool — it's a conversation about who defines the data priorities that marketing depends on. Without that, every AI use case requires permission from another team to progress.
Have the highest-value use cases been defined? Content generation is not a strategy. The use cases that change what a marketing function can know and decide — performance analysis, audience activation, attribution, real-time optimisation — need to be named, sequenced, and resourced before tools are selected.
Is there a measurement framework that can evaluate AI investment? Not a perfect one. A directional one. If the only way to justify AI spend is a precise ROI calculation that marketing data can't currently support, the investment will always be vulnerable. A signal-based approach — connecting AI adoption to the decisions it improves and the outcomes those decisions affect — is more honest and more defensible.
The 94% who haven't fully embedded AI aren't failing because the technology isn't good enough. They're failing because the conditions for AI to work — data foundations, strategic ownership, defined use cases — haven't been built. That's a leadership problem before it's a technology problem.
Where to go deeper
The full data is in the Supermetrics 2026 Marketing Data Report, which surveyed 435 marketers globally and covers AI adoption, data ownership, measurement maturity, and activation gaps in detail.
For the external dimension of this problem — how AI adoption gaps affect how your brand appears to buyers using AI tools to research — my post on LLMs and brand visibility covers the other side of the same challenge.
Andrea Linehan is CMO at Supermetrics, a B2B marketing intelligence platform used by 200,000+ marketers globally. She is a four-time CMO across MarTech and FinTech, and writes about AI, marketing strategy, and the finance discipline that connects them.
Frequently asked questions
Why is AI adoption in marketing so low?
According to the Supermetrics 2026 Marketing Data Report, which surveyed 435 marketers globally, only 6% have fully embedded AI into their workflows despite 80% being under pressure to do so. The primary barriers are a lack of clear strategy from leadership, fragmented data foundations that can't support AI use cases, and the absence of marketing involvement in data strategy decisions. Most AI experiments are happening in silos without defined use cases or the infrastructure to scale them.
What are the most successful marketing teams doing differently with AI?
Based on the 2026 Marketing Data Report findings, the marketers who have fully embedded AI into their workflows share four characteristics: they built clean, connected data foundations before selecting tools; they defined specific business use cases before deploying technology; marketing leadership is meaningfully involved in data strategy; and they treat ROI as a directional signal rather than a precise calculation. The sequence — foundation first, tools second — is the most consistent differentiator.
Why does data ownership matter for AI adoption in marketing?
When data strategy is owned by external teams, marketing loses control over the priorities, definitions, and governance that determine what data is available and how it can be used. The 2026 report found that 52% of marketers say data strategy decisions are made outside marketing, and only 31% of CMOs are meaningfully involved. Without that involvement, every AI use case requires cross-functional permission to progress — which is why adoption stalls even when the tools and the budget are available.
What is marketing AI actually being used for — and what should it be used for?
The 2026 Marketing Data Report found that 87% of marketers are using AI for content creation and copywriting — the lowest-leverage application available. Only 39% are using it for reporting and analytics, and 33% for marketing automation. The use cases most likely to change what a marketing function can know and decide — performance analysis, audience activation, real-time optimisation — are the least commonly adopted, and represent the largest unrealised opportunity.
How should CMOs make the case for AI investment at board level?
The most defensible approach is signal-based rather than precision-based. Rather than attempting to calculate exact ROI from AI investment — which current marketing data rarely supports — CMOs should connect AI adoption to the decisions it improves and the outcomes those decisions affect. The 2026 Marketing Data Report notes that 40% of marketers struggle to prove ROI across channels, which suggests the measurement framework needs to be built alongside the AI strategy, not after it.