Agents
Build patterns, frameworks and cautionary tales for AI agents and automations that do marketing work, from a practitioner who runs them in production.
Agents are the fastest moving part of this stack and the most oversold. This section covers what an agent actually is in a marketing context: software that takes a goal, plans steps, calls tools, and delivers a result. It covers the patterns that survive contact with real campaigns: prompt pipelines, human-in-the-loop review gates, tool call budgets, error handling, and where a cron job beats an agent every time. I write from builds I have shipped, and when something comes from the docs or a demo rather than my own run, it is marked as such. Expect workflow walkthroughs, framework comparisons, cost breakdowns, and honest notes on failure modes, because most agent projects fail on the boring parts, not the model.
- Building marketing agents with Make.com: what we learned
We built marketing agents in Make.com and learned where the canvas helps, where the credit meter bites, and what a real stack costs in 2026. Dated and sourced.
- Building marketing agents with MCP servers: what we learned
We built marketing agents on MCP servers. Here is what held up, what failed, and the approval gates that saved us, dated August 2026.
- Building marketing agents with n8n: what we learned
Building marketing agents with n8n: what held up, what failed, and what a stack costs in 2026. Dated and sourced, before you build.