Agentic AI in marketing
Where an agent adds judgement and where a well-built automation is enough.
Notes from a CMO on agentic AI, on how marketing is organised and on how B2B companies grow.
Topics
What I write comes from the daily work of running marketing at a B2B company growing across several markets.
Where an agent adds judgement and where a well-built automation is enough.
Every process has an owner and a checkpoint
Processes, teams and what a person always decides, even when agents do the rest.
Target accounts, pipeline and working with sales, market by market.
How I work
All three lines run through the same point: the moment a person decides. Pick a line to see its route.
The trains never skip the central station: that is where a person reviews, prioritises or changes course.
Agents research and draft. Automation repeats what is defined. What reaches the customer is decided by a person.
Who decides what
Before adding an agent, know which layer it works in and who answers for what it does.
Self-assessment
Ten questions on strategy, data, processes, team and measurement. At the end you will see your team’s agentic maturity and where to start.
About 3 minutes. Your answers never leave your browser.
Notes
I am a CMO at a B2B SaaS scale-up in Madrid. For ten years I have built marketing teams and systems, from consumer apps to selling to large accounts across Europe and LATAM. Here I write about how marketing work changes when agents do part of it, and about what a person must keep deciding.