Course
AI for Digital Marketing (Hard)
Twelve accounts, two juniors, every grind automated and every decision still yours.
Start with
- AI at Work: FoundationsStart from zero and finish with three things you can actually use.
- AI for Digital Marketing (Beginner)Your first year running social, reviews and a small ad budget alone.
Who this is for
Agency account leads and senior in-house marketers who run many accounts and want the grind automated and the judgment kept. Every example follows Nour, an account lead at a small agency running Google and Meta ads for twelve local clients, with two juniors.
You will be able to
- Connect Google Ads and GA4 through MCP and audit any account in ten minutes
- Run the weekly search-term and placement cleanup as a Cowork delegation you review, never an auto-apply
- Build the twelve-client report pipeline with Claude Code and ship it monthly
- Quality-control the juniors campaign builds with a checklist Claude runs first
- Write the client review that leads to a renewal and a budget conversation
- Define the guardrails: what Claude may recommend, what only a human changes
Stops
- Connectors and the ten-minute auditWire Google Ads and GA4 to Claude, then get a ranked findings table for any account in ten minutes, read only. Set once
- GuardrailsThe written rule set that decides what Claude may recommend and what only a human changes, plus the approval message and the log. Set once
- Search terms and placements with CoworkDelegate the weekly cleanup across twelve accounts, review the proposal against the never-negate list, and apply the survivors by hand. Weekly
- Performance Max diagnosisAsset groups, search categories and the brand-traffic question, plus the three changes worth proposing and the ones the data cannot support. Weekly
- Meta: creative fatigue and the test planRead fatigue out of the Meta export, size a test plan that can actually resolve, and brief the designer in one page. Weekly
- The report pipeline with Claude CodeDescribe the twelve-client report job once, let Claude Code build the script, and put a QA step in front of every run. Weekly
- Quality control for juniorsA build checklist Claude runs before Nour looks, feedback the juniors can act on, and a defect count that should fall. Daily
- The client reviewTurn twelve numbers into one story, attach a confidence line to the recommendation, and make the budget conversation part of the review. Weekly
- Nour's month and the capstoneThe three scheduled routines that carry the month, then the graded capstone: one account from audit to review with its guardrail log. Weekly

Written for Claude. The ideas transfer to ChatGPT and Gemini; the menu paths don't. Current as of September 2026; where a feature may change we link to support.claude.com.
Claude helped Dasow draft and structure every course. A human at Dasow reviewed every stop.