A new survey of senior marketing leaders just confirmed something a lot of business owners already suspect: marketing departments are being asked to do more with AI, while getting less support to actually learn how to use it well.
Duke University's Fuqua School of Business runs an ongoing poll of marketing chiefs, and the latest edition, completed in January 2026 with 308 senior marketing leaders, found training budgets have fallen to 3.8% of marketing spend. Before the pandemic, in 2019, that number was 5.8%. Headcount growth in marketing teams has also dropped by more than half compared to last year.
At the same time, AI use inside marketing departments has nearly doubled in two years, going from about 13% of marketing tasks to roughly 24% today. Marketing leaders expect AI to handle more than half of all marketing work within three years. So the tools are arriving fast, but the people trained to run them well are not keeping up.
This is not just an American problem or one survey's fluke. A separate study by Gartner, covering 401 marketing chiefs across the US, UK, and Europe, found the identical pattern: most want their company to be seen as an AI leader, but only about three in ten feel their internal processes are actually mature enough to make that happen. More than half say they lack the right talent, and a similar share say they lack the budget too.
Why does this keep happening? Two reasons stand out. First, the relationship between marketing leaders and the finance chiefs who approve budgets is weak and has barely improved in years. Second, marketing leaders tend to justify training mainly by pointing to short-term returns, rather than to harder-to-measure benefits like being difficult for competitors to copy or holding on to good staff. When a budget can only be defended with short-term numbers, it loses that argument the moment profits dip, and marketing spending gets cut before almost any other department's budget does.
There's a real cost to this. New skill categories are appearing that did not exist two years ago, generative engine optimization being one example: the job of making sure your brand actually gets mentioned when someone asks ChatGPT, Google, or Perplexity for a recommendation instead of a competitor. If nobody on your team is watching that, you become invisible in a growing share of customer research, without ever knowing it happened.
Meanwhile, smaller companies appear to be adapting faster in some ways. More than half of small businesses already use AI marketing tools, and adoption is expected to keep climbing this year. They don't have layers of budget approval to fight through, they just try the tool.
The lesson for any business that depends on a marketing team, whether that team is three people or three hundred, is to stop treating training as the first thing to cut when money is tight. Treat it more like research spending: cut it and you save money this quarter, but you fall further behind competitors who kept investing, and that gap is much harder to close later than it was to avoid in the first place.