AI in Marketing: Where Marketing Budgets Are Shifting

In short: AI in marketing is shifting budgets. A growing share goes to technology and in-house teams, while agencies stay strong in strategy and brand. For budget planning, that means re-scoping agency work and deciding up front which metric should show the effect of the shift.

Within just a few years, marketing budgets have shifted noticeably, and AI in marketing is one of the main reasons. A growing share flows into technology and in-house teams, while a smaller share stays with outside agencies. The shift follows a clear logic, even if it rarely gets called out in budget discussions.

In conversations with other CMOs, I keep hearing the same order of magnitude. Three years ago, the bulk of the operating budget went to agencies. Today, a noticeable portion is moving to in-house teams and to system licenses. That doesn’t automatically mean less total budget for outside partners, but it does mean a different kind of brief.

Where agencies are strong today

Agencies remain strong where strategy, creativity and brand management are at stake, in other words where questions call for deep judgment and experience across many industries. That work is hard to scale, which is exactly why it keeps its value. I still work with agencies myself, but the briefs have shifted: less production, more advice on brand questions that no system can answer.

What in-house teams can do today

At the same time, small in-house teams can now handle work that used to require large outside capacity, from content production to ongoing reporting. Systems take on much of the execution, while the team steers and owns the result. A team that once needed weeks of outside support for a major campaign can now handle a large part of it on its own, because the tools have become accessible.

What this means for budget planning

For budget planning, it pays to split budgets along one question: where is judgment actually needed, and where is it execution? I’ve started building exactly this distinction into our own budget planning, and it makes conversations with the CFO much more concrete, because every line item can be tied to a clear rationale.

A new division of labor

The result is a new division of labor: outside partners for the questions that need experience and an outside perspective, in-house teams for the tasks that are closely tied to day-to-day business and require fast iteration. Those who design this split deliberately gain speed in daily work and substance on the big questions, and avoid the situation where both sides work on the same task in parallel without knowing it.

In the end, budgets always follow the question of where the greatest value is created.

That is exactly why it pays to shape this shift actively, with a clear stance on what should be built in-house and what deliberately stays outside.

What should the budget for AI in marketing be measured against?

Today, the success of AI in marketing is mostly measured in time saved. Hardly anyone is truly satisfied with that, because saved hours rarely show up in the leadership team’s report. The IBM Institute for Business Value found that the best-performing companies are 21 percent more likely than others to tie the results of their organizational change explicitly to financial metrics.

For budget planning, this means: if you shift money from agencies to licenses and in-house teams, decide up front which number should show the effect. That could be a campaign’s cycle time or the pipeline from one channel. Saved hours alone aren’t enough.

A second number belongs next to it. According to IBM, 81 percent of executives say their companies reward employees for building AI skills. Only 64 percent reward those who question AI outputs. Usage gets rewarded, and judgment gets lost. That’s why I treat it as a fixed rule to pair every speed metric with a quality counter-metric. If you track how many texts go out without edits, also track how many of them had to be corrected later.

What should leadership teams review now?

  • One question: Which tasks our agencies handle today could we now do ourselves with AI in marketing, and which should deliberately stay external?
  • One metric: the share of the marketing budget spent on technology and in-house capability compared with outside providers, tracked over three years.
  • One conversation: with our most important agencies about what their role should look like in two years.

Frequently asked questions

How is AI in marketing changing budget allocation?

Production and routine analysis are getting cheaper and are moving in-house more often. Budget shifts toward technology, data and the people who can run both.

Will companies still need agencies?

Yes, for work that calls for exceptional creativity, market knowledge or capacity for peaks. The collaboration becomes more targeted, and agencies are measured more on results.

Which marketing tasks are best suited for AI first?

High-volume tasks with clear quality criteria, such as ad and copy variants, translations, campaign reporting or maintaining product information. For brand management, positioning and crisis communication, human judgment remains decisive.

What does the shift mean for the marketing team?

The team needs new skills in-house: people who can run AI tools, review output and analyze data. Anyone moving budget from agencies in-house has to build that capability at the same time. Otherwise a gap opens up that first shows in quality.

Sources: IBM Institute for Business Value (2026): Where AI breaks or breaks through

Similar Posts