AI agents for marketing teams
Marketing teams produce the highest volume of structured, repeatable content work — research, drafting, enrichment, reporting. These guides cover what workflows to build as agents, how to design the human gates that protect brand voice, and where to start.
Start here
AI agents for marketing teams: what to build and where to start
Five marketing workflows worth building as AI agents — content research, social drafting, lead enrichment, campaign reporting, and SEO monitoring. With human review gates designed in from the start.
HubSpot AI agents: what to build and where to start
Lead enrichment, contact scoring, deal monitoring, and CRM automation — the HubSpot workflows that run well as AI agents, and how to design them safely.
Human-in-the-loop: how to design approval gates into a multi-agent workflow
Brand voice, compliance language, and public-facing content all require human review. Here's how to design gates that protect quality without breaking the workflow.
How to design AI agents: a practical guide
The six design decisions that determine whether an AI agent system works — applicable to every marketing workflow from content production to campaign reporting.
AI agent costs: how to estimate and control your spend
Model inference, tool calls, and human gate time — the three cost components of any agent workflow, and the levers that bring spend down without reducing quality.
Sub-agent design: the five components every agent needs
Role, skills, model, tools, and handoffs — how to define each component clearly before you build, with examples from content and enrichment workflows.