The guide, 18 chapters

AI agents.

Agents that do real growth and prospecting work, plugged into your tools, with a check before anything goes out.

Build my agents

Introduction

What an agent is, what it can do today, and which tools to use.

  1. 01Why AI agents change B2B growthWhat moves when an agent does the repetitive work of B2B growth, what still needs you, and where to start in your own team.
  2. 02AI agent, automation, chatbot: the differencesThree things sold under the same word, and when to use each one.
  3. 03What an agent can do today, and what it still gets wrongWhere agents are reliable, where they fail, and how to tell before it costs you.
  4. 04Choosing your tools: Claude, Claude Code, Make, n8n, scriptsWhich tool for which job, from a no-code flow to an agent that writes code.

The steps

From an existing process to an agent that runs on its own and is measured.

  1. 05Start from a process that already existsPick the task, write down how it is done today, and define what done looks like.
  2. 06Write an agent’s role, rules and examplesThe instructions file that turns a model into a reliable colleague.
  3. 07Connect the agent to your tools with MCPCRM, spreadsheets, databases and APIs, without copy and paste.
  4. 08Build your own MCP serverWhen an existing connector is not enough, and how to design the tools you expose.
  5. 09Give the agent memory and contextWhat an AI agent must know every time, what it can look up when needed, and where each fact should live so it stays true.
  6. 10Put a quality check before anything goes outMechanical checks, an independent judge and frozen test cases.
  7. 11Run the agent on its ownScheduled routines that check their tools first and propose before they act.
  8. 12Measure what the agent really brings inOutput against outcome, dated verdicts and instruments you can trust.
  9. 13Security, customer data and GDPRSecrets, personal data, permissions and GDPR: how to set up an AI agent safely, and the actions it must never take alone.

Use cases

Prospecting, programmatic SEO, LinkedIn and a full marketing team.

  1. 14A B2B prospecting agentTargeting, enrichment, signal and message, with a human on the send button.
  2. 15A programmatic SEO agentPages generated from data, checked before they ship and tested after.
  3. 16A LinkedIn agent that writes in your voiceVoice calibration, a weekly cadence and a final human pass.
  4. 17A team of agents for all of marketingHow Derrick’s marketing runs on Claude agents, and what broke along the way.
  5. 18AI agents or a part-time head of growthWhat to hand to agents, what to keep for a person, and how to split the two.