What 3,200 Leaders Told Deloitte About AI, and What Marketing Leaders Should Hear.

The Access Assumption

The logic behind most AI rollouts follows a clean sequence. Evaluate tools. Select vendors. Provision access. Train the team. Measure adoption. The assumption embedded in that sequence is that access leads to usage, and usage leads to value.

The Deloitte data says otherwise. Access grew dramatically. Usage didn't follow. And the report identifies the reason: most companies are focused on building AI fluency (educating workers on how to use the tools) rather than redesigning the work itself around what AI can do.

For marketing leaders, this maps to a familiar pattern. The team gets access to an AI writing tool. Someone runs a training session. A few people start using it for first drafts. And then the tool sits alongside the existing workflow, generating output that still needs to be edited, reconciled, and quality-checked through the same manual process that existed before the AI showed up.

The workflow didn't necessarily change, but the tools added a step at the beginning of it.

What the Other Eighty-Four Percent Look Like

Deloitte breaks companies into three groups based on how deeply they're integrating AI into their operations. Thirty-four percent are using AI to deeply transform their businesses, creating new products, reinventing core processes, or fundamentally changing their business models. Thirty percent are redesigning key processes around AI while keeping their business models intact. And the remaining 37 percent are using AI at a surface level, with little or no change to existing processes.

All three groups are capturing productivity gains. Only the first group is actually reimagining how the work gets done.

On a marketing team, surface-level AI adoption looks like this: individual contributors use AI to generate first drafts. Each person prompts differently. Each person applies a different interpretation of the campaign brief. The outputs arrive faster, but they arrive disconnected from each other, disconnected from the positioning, and disconnected from the conversion path the campaign was designed to support. Someone has to reconcile all of it manually, and that reconciliation is where the time savings disappear.

That's from a healthcare AI leader interviewed for the Deloitte report, describing what happens when organizations fund experiments without building a path to production. The same dynamic plays out in marketing.

Teams launch AI pilots, see promising results in controlled conditions, and then can't figure out how to make those results repeatable across a full campaign cycle.

Why Education Alone Doesn't Close the Gap

The Deloitte report found that 53 percent of companies are focused on educating their broader workforce to raise AI fluency. Only 33 percent are redesigning career paths and career mobility strategies. Only 30 percent are reimagining organizations based on new work patterns resulting from AI usage.

Education is necessary. It's also insufficient. Teaching someone how to write better prompts doesn't solve the structural problem. The structural problem is that campaign planning, content production, and launch coordination were designed for a pre-AI workflow, and nobody has redesigned those processes to account for what AI actually produces and what it doesn't.

Consider what a marketing manager's week looks like when AI fluency is the only intervention. They know how to prompt effectively. They generate drafts quickly. Then they spend the rest of their time doing the same work they did before: manually checking every draft against the campaign brief, manually ensuring consistency across assets, manually verifying that proof points are accurate and CTAs are aligned. The AI made the first five minutes faster. It didn't change the next five hours.

The Deloitte data supports this at an enterprise level. Seventy-four percent of organizations hope to grow revenue through AI initiatives. Only 20 percent are actually doing so. The gap between aspiration and result is the gap between giving people tools and changing the system those people operate inside.

The Redesign Question

The 16 percent of companies that have redesigned work around AI share a common trait. They didn't start by asking "which AI tools should we deploy?" They started by asking "how does the work need to change?"

For marketing teams, that question gets specific fast.

• How should a campaign brief be structured so AI can produce assets that stay on-message without constant human correction?

• What needs to be defined upstream (segments, positioning, conversion paths) so downstream production doesn't require individual judgment calls at every step?

• Where should human expertise concentrate, and where should the system handle the repetitive structural work?

These are design questions, not technology questions. The answers determine whether AI tools sit on top of an existing process (adding speed at the beginning and rework at the end) or operate inside a process that was built to take advantage of what they produce.

Most marketing teams are in the 84 percent right now. They have the tools, the access, and even have team members who are genuinely skilled at using AI. What they don't have is a campaign production system that was designed for this era of work.

The Deloitte report frames the core challenge as moving from "ambition to activation." For marketing leaders, activation doesn't mean more AI training or another tool evaluation. It means redesigning how campaigns get planned and produced so that AI operates inside a structure, rather than alongside one that was built without it.