CMOs Say AI Is Their Top Priority. Their Budgets Say Otherwise.

Eight out of ten CMOs told Gartner that AI is their number one strategic priority for 2026. Then Gartner looked at the budgets. AI accounts for 8 to 10 percent of direct marketing spend. That's it.

There's a gap here. A big one. And it tells us something important about how marketing organizations actually adopt technology versus how they talk about it.

The Numbers Don't Add Up (Until They Do)

It's not that CMOs are lying. It's that most marketing AI spend is invisible. It's buried inside existing line items. Your agency raised its rates 12 percent last year? Part of that covers the AI tools they're using to produce your creative faster. Your content production costs went up? That line item now includes AI-generated first drafts and synthetic image licensing.

CMOs are spending on AI. They just can't see it. And what you can't see, you can't measure.

The Attribution Problem

Marketing has spent two decades building attribution models. Multi-touch, last-click, algorithmic, you name it. Those models were designed to answer one question: which dollar drove which outcome? AI breaks that question.

When an AI tool rewrites your email subject lines and open rates climb 15 percent, the attribution model credits the email channel. Not the AI. The result: CMOs know AI is working. They feel it. But the dashboards don't reflect it. The ROI case stays fuzzy. And when budget season comes around, fuzzy doesn't win arguments.

Spending More vs. Spending Differently

The companies getting real results from AI marketing aren't the ones spending the most on it. They're the ones spending differently. The difference comes down to a simple question: are you adding AI on top of existing workflows, or are you replacing workflows entirely?

Most marketing teams are doing the former. They buy an AI writing assistant. Their writers use it for first drafts. The drafts still go through the same review cycle, the same approval chain. Net time saved: maybe 20 percent. The teams pulling ahead took a different approach. They looked at a workflow end to end and asked what it would look like if they designed it from scratch with AI at the center.