Founders who just hired an AI CMO on Tycoon (or any AI marketing role) and want a disciplined onboarding track. Also useful as a reference for onboarding any AI specialist — the pattern generalizes.
First production-quality deliverable on day 4-5. Full autonomy on routine work by day 14.
The playbook
- 1
Day 1 — Context dump, not delegation
Don't ask the CMO to produce anything yet. Spend 60-90 minutes loading context: your positioning doc (or rough notes if none exists), your last 5 pieces of marketing content for voice sample, your customer avatar (real names and quotes from 3-5 customers), your competitor list with strong/weak assessment, your goals for the next quarter. If any of this doesn't exist in writing, this is the week to write it. The CMO's output quality is a direct function of context quality.
Notion or Google Docs for the positioning brain dumpLoom for a 20-minute founder narrative of the business - 2
Day 2 — Voice calibration
Have the CMO draft three pieces in your voice: a landing page headline, a social media post, a customer email. Critique them like a senior editor — specific, blunt, example-driven. 'This reads too corporate; we never say leverage; make the second sentence shorter.' The CMO updates its voice guide after each correction. After 6-10 rounds of correction on day 2 the voice is usually 80% right.
Your own tasteThe CMO's draft output in chat - 3
Day 3 — Analytics and audit
Give the CMO access to GSC, GA4, PostHog (or your equivalent), and your social accounts. Have it produce a one-page marketing state-of-the-business: what's working, what's leaking, what looks promising, what looks dead. This is a diagnostic, not a plan — plans come later. The point is to validate that the CMO is reading the numbers correctly.
GSC read accessGA4 read accessPostHog or Mixpanel API keyBuffer/Typefully/social platform access - 4
Day 4 — First production deliverable
Have the CMO ship one real thing: a blog post, a landing page, a newsletter, a launch brief. You review before publish. Quality should be noticeably higher than a cold-AI output because of days 1-3 context work. If quality is not there, the gap is almost always context, not capability — go back to days 1-2 and fix. Most founders are surprised by how good day 4 output is when days 1-3 are done properly.
Your CMS or blog publishing toolYour ability to review with senior taste - 5
Day 5 — Cadence and scheduled tasks
Define the weekly cadence: Monday content calendar, Wednesday newsletter, Friday roll-up. Define the monthly cadence: positioning review, channel rebalance, content refresh. Define what the CMO escalates to you (major positioning changes, significant spend decisions, brand voice corrections) and what it decides alone (ordinary blog topics, social schedule, small A/B tests). Write these down — verbal agreements get lost.
Notion or a simple Markdown doc for the cadence specTycoon's autonomy settings per role - 6
Day 6-7 — Ship the week's actual work
By day 6 the CMO should be shipping real deliverables at the cadence you defined. For most B2B SaaS that's: 2-3 content pieces, 1 landing page, 1 newsletter, 5-10 social posts, 1 analytics review. You review for voice and strategy, not for the basic execution. If you're still editing grammar and word choice on day 7, go back to day 2.
Your review queueThe CMO's output pipeline
Pitfalls to avoid
- Skipping day 1 and starting with day 4 — you'll get generic output and blame the AI.
- Critiquing output with 'make it better' — useless. Always be specific: 'shorter', 'less corporate', 'start with the objection'.
- Not writing the voice corrections down — the CMO's memory compounds only if you're consistent.
- Delegating launch day in week 1 — wait until week 3-4 for high-stakes work while trust compounds.
- Expecting the CMO to know your positioning without you writing it down — it can infer only what exists.