How CMOs are using
AI to Change Marketing
100 CMOs across Tech, FinServ, Healthcare, Consulting, and other on how AI moved from experiment to operating system.
01
Four shifts redefining the CMO's stack
AI has crossed from experimentation into operations. Across 100 CMOs, four patterns stood out, and together they change how brands get discovered, chosen, and built.
AI has become need-to-have for the CMO
90% of CMOs use LLMs daily. The 10% who haven't shifted are now the exception.
CMOs are building model stacks
Work is routed to the best model per task. Claude leads strategy and long-form at 44%.
LLMs are forming the consideration set
90% of CMOs use LLMs daily. The 10% who haven't shifted are now the exception.
Systems over prompts
90% of CMOs use LLMs daily. The 10% who haven't shifted are now the exception.
02
AI moved from nice-to-have to need-to-have for the office of the CMO
For most CMOs, AI has stopped being a place to experiment. It is simply where the work gets done now.
Use LLMs daily
Daily use is now the default operating mode for marketing leaders.
Say usage climbed in the past 6 to 12 months
Adoption isn't leveling off. It's still accelerating inside marketing teams.
Haven't made the shift
The holdouts are now the exception rather than the norm.
03
CMOs are building model stacks and routing work across them
Leaders send each task to whichever model handles it best.
Task-by-platform preferences
Claude for long-form & strategy
ChatGPT for content & copy
Gemini for Google ecosystem & data
Claude for analysis & research
No strong preference
Copilot for Microsoft & enterprise
ChatGPT for brainstorming
Perplexity for research & citations
Open-ended responses on which model CMOs prefer for which task (Q14) · N=100 · multi-label, bars do not sum to 100%
"I'm a bigger fan of Claude from a strategic standpoint. It's almost like seeing the forest for the trees, whereas ChatGPT feels a bit more transactional to me."
CMO Respondent · Open-Ended, Q14
93% of CMOs have at least moderate trust in their preferred LLM for business-critical marketing work. Trust is driven by output quality and reliability (37%), familiarity and track record (31%), and security and privacy controls (17%).
Task-by-platform preferences
Claude for long-form & strategy
44%
ChatGPT for content & copy
29%
Gemini for Google ecosystem & data
23%
Claude for analysis & research
21%
No strong preference
14%
Copilot for Microsoft & enterprise
12%
ChatGPT for brainstorming
11%
Perplexity for research & citations
6%
Open-ended responses on which model CMOs prefer for which task (Q14) · N=100 · multi-label, bars do not sum to 100%
"I'm a bigger fan of Claude from a strategic standpoint. It's almost like seeing the forest for the trees, whereas ChatGPT feels a bit more transactional to me."
CMO Respondent · Open-Ended, Q14
04
LLMs now beat Google as the first stop for the CMO
When CMOs size up a new tool, an LLM is now the second place they go, ahead of Google and behind only other people. Brands have to worry about more than where they rank in search. What counts now is how credibly they show up in AI-generated answers.
Where CMOs start research on a new tool
Peer recommendations & networks
49%
An LLM (ChatGPT, Perplexity...)
22%
Google or another search engine
23%
Industry publications & analysts
21%
Internal company resources
14%
Other
12%
Starting point when evaluating new software or services (Q26) · N=99 · single-select
"I'll head to the LLM first, get its hot take on my question, then go back to a traditional search engine to pull more, or follow the links the LLM gave me."
CMO Respondent · Open-Ended, Q26
05
The advantage is moving from prompts to systems
Nearly half (47%) have moved beyond ad hoc prompting into structured AI workflows, automated reporting, and custom agents. Compared with other marketers, power users are far more likely to use AI for customer insights and segmentation (74% vs. 20%), campaign strategy and planning (82% vs. 40%), and data interpretation and reporting (84% vs. 50%).
06
AI is changing how marketing teams are built
AI is reshaping organizations as well as workflows. 28% say AI fluency is now a hiring requirement, while 32% expect AI to reshape team structures or reduce marketing headcount.
Run structured AI workflows & agents
Automated reporting, orchestration, and custom agents, well past one-off prompts
Require AI fluency when hiring
AI skill is becoming a baseline expectation for the modern marketer.
Expect team restructuring or smaller teams
The shape of the marketing org is being redrawn around AI.
Three moves for the AI-native CMO
Nearly half of CMOs have moved past ad hoc prompting into structured workflows, automated reporting, and custom agents. As AI becomes part of core marketing systems, the organization is adapting around it.
01
Build for specialization
Instead of standardizing on one model, send work to whichever model fits each task and each budget best.
02
Use LLM visibility as a growth channel
As buyers turn to LLMs to evaluate vendors, optimize how your brand is represented in AI-generated answers.
03
Build AI-native teams and workflows
Hire for AI fluency and rebuild core workflows around AI instead of leaning on one-off prompts
AI-led User Interviews
Use Case