Seventy percent of retail, e-commerce, and CPG marketers say optimizing ad spend is their top priority. Only 17% use AI to actually do it. That’s the central finding from the latest Supermetrics Marketing Data Report, released on March 24, and it exposes a gap that most marketing teams haven’t figured out how to close.
The report surveyed over 400 marketing professionals globally. Campaign optimization turned out to be the least-adopted AI use case across all categories studied, sitting well below content creation (38%) and workflow automation (27%). The disconnect is striking: the thing marketers care about most is the thing they’re least likely to use AI for.
The Numbers by Industry
The adoption gap varies significantly by sector. Retail marketers lead with 22% using AI for campaign optimization. CPG and FMCG teams follow at 14%. E-commerce brands, despite operating in one of the most data-rich environments in marketing, sit at just 8%.
That last number is the surprising one. E-commerce companies generate real-time conversion data, basket-level purchase signals, and granular customer behavior data across every touchpoint. They have more raw material for AI-powered optimization than almost any other category. Yet they trail retailers by nearly three-to-one in actually applying AI to their campaign decisions.
The report points to three primary blockers. Thirty-eight percent of respondents cite a lack of in-house expertise. Thirty percent say their technical infrastructure can’t support AI workloads. And 27% can’t clearly articulate the business value or ROI of AI implementation. In other words, it’s not that teams don’t want to use AI for optimization. They don’t have the people, the systems, or the proof points to justify the investment. The May 11 Gartner CMO Spend Survey readiness gap measures the same blockers at the CMO level: 70% of marketing organizations say they don’t yet have the process maturity to scale the AI tools they’ve already bought.
The Bigger Picture: AI Adoption Is Stalling at 6%
This latest data builds on Supermetrics’ earlier finding from February 2026. That report, based on a survey of 435 marketing leaders, found that 80% of marketers feel pressure to adopt AI, with 89% of that pressure coming from the C-suite and board. Despite that top-down urgency, only 6% have fully embedded AI into their workflows. McKinsey’s State of Marketing Europe 2026 report independently confirmed this exact figure — 94% of marketing teams are stuck at low or moderate maturity. Gartner’s May 2026 automation forecast sharpens the deadline: the 16% baseline marketing leaders self-report needs to reach 36% by 2028, a 20-point swing in 30 months that lands hardest on the 94% still building foundations.
The gap between executive expectation and team reality is one of the defining challenges for marketing operations in 2026. Leadership sees the headlines about AI agents transforming revenue operations and wants results. Marketing teams see fragmented data across 12 platforms and a three-day wait for data team support. The B2B CMO Project’s Imperative report names AI visibility as one of five 2026 imperatives, putting the operator-side translation directly on the marketing team’s plate. The GrowthLoop 2026 Index puts a number on the operational cost of an unprepared measurement layer: 77% of winning A/B tests fail at scale, which is what happens when adoption widens before the causal-data spine underneath catches up. Canva’s 2026 AI marketing trust report shows the creative version of the same gap: output can scale before quality control, disclosure, and brand review are ready.
More than half (52%) of respondents said external teams, not marketing, define their data strategy and measurement framework. When the people closest to campaign performance don’t control how data gets collected, structured, and activated, AI tools can’t access the clean inputs they need. You end up with AI experiments running in silos, driven by excitement rather than a connected data model. eClerx’s 78% activation-gap finding measures the same failure one step later: the insight can exist, but partial data and slow workflows still prevent the team from acting on it.
What This Means for B2B Teams
B2B marketers face the same structural problem, often amplified by longer buying cycles and smaller datasets. When you’re running account-based campaigns with deal cycles of three to nine months, the feedback loop for AI optimization is inherently slower than in e-commerce. That makes data quality even more important. Every signal you feed the model needs to be accurate because you don’t have millions of transactions to smooth out the noise. DemandScience’s April 2026 intent-data research shows a similar pattern in B2B account selection.
The contrast with the sales side is revealing. Salesforce’s 2026 State of Sales report shows 87% of sales teams now use AI — far ahead of marketing. Sales teams have clearer data models (CRM records, deal stages, win/loss) that make AI optimization more straightforward. Marketing’s data is messier and more fragmented, which is exactly why RevOps alignment matters.
And the stakes are rising: Gartner predicts 40% of agentic AI projects will be canceled by 2027, primarily due to poor data foundations. Teams that rush into AI tools without fixing their data infrastructure first are the ones that end up in that 40%. The SaaS-procurement side is consolidating into the same workforce platforms B2B operators already use: the Sastrify-Deel IT consolidation on May 5 folds renewal management, usage intelligence, and a $2B benchmarked-contract dataset into Deel IT, which raises the floor on what marketing teams should expect from their own procurement audit before stacking the next AI tool on top.
The practical takeaway from the Supermetrics data isn’t to rush into AI adoption. It’s to fix the prerequisites first. That means three things. First, audit your data connections. Can you link ad platform data, CRM records, and revenue outcomes in a single view without manual spreadsheet work? If not, that’s your first project. Second, take ownership of the measurement framework within marketing instead of delegating it to IT or a data team that optimizes for collection, not action. Third, start with one high-value use case for AI rather than trying to automate everything at once. Campaign budget allocation is a good candidate because it has clear inputs, measurable outputs, and immediate financial impact.
The Supermetrics report is available as a free download on their website. It includes a nine-step AI readiness checklist that’s worth running through before your next planning cycle, regardless of which tools you’re evaluating.
