A Playbook That Teaches CMOs to Lead AI
We just published a new playbook template in T2D3 OS: CMO Mastery — Leading T2D3 OS. Seven modules and twenty-nine tasks that teach the one skill the AI era actually demands of marketing leaders: pairing probabilistic tools that draft, rank, and recommend with the deterministic leadership only a human can provide. Here is why we built a curriculum instead of another feature, what "the machine proposes, you decide" looks like as a weekly practice, and why the capstone is you publishing your own point of view.
Stijn Hendrikse · Aug 28, 2026
Most AI marketing tools ship features and hope you figure out the operating model. We just did the opposite: we shipped the operating model as a product. CMO Mastery: Leading T2D3 OS is a new playbook template in our library — seven modules, twenty-nine tasks — and its subject is not the tool. Its subject is you.
The skill nobody is teaching
Here is the uncomfortable truth about AI in marketing: the models are no longer the bottleneck. The leadership is.
An AI system is a probabilistic machine. It drafts, ranks, estimates, and recommends — fluently, tirelessly, and with initiative. But fluency is not truth, and a recommendation is not a decision. What a marketing org needs around that machine is something entirely deterministic: a locked ICP that downstream work can trust. A brand standard that is law, not a suggestion. A vote that was cast, a number that was confirmed, a decision with a name and a date on it.
Most CMOs I talk to make one of two mistakes. They edit every AI sentence — competing with the machine at the thing it does best. Or they accept every AI suggestion — quietly resigning from the thing only they can do. Both waste the pairing. The playbook exists to train the third way: the machine proposes, you decide.
What the curriculum actually covers
The seven modules walk the pairing through every surface of a marketing org:
- The operating model — learn to look at any screen and know whether the machine is proposing or a human is deciding. Tour the lock gate. Learn what we call glint: the novel insight only you can contribute, the scarcest input in the whole system.
- The foundation — interrogate the AI's ICP draft the way a board member would, then lock it with written judgment. Pressure-test value props against the claims your competitor could copy-paste. Lock brand so the machine can scale it.
- Signal mastery — feed the machine ten curated documents instead of a hundred dumped ones, then practice feedback as a management discipline: rule-shaped corrections, verified on the next generation.
- Leading the four roles — Navigator, Scribe, Sculptor, Engineer: the AI org chart. Set autonomy ceilings deliberately, write a delegation charter, and run a full week of intent-based leadership — straight from David Marquet's Turn the Ship Around! — where the OS leads and you only steer.
- The content machine — walk one article from calendar to channel by hand, define your publish bar as three written checks, audit your content for answer-engine legibility, and wire auto-promotion so distribution is machinery, not memory.
- Measurement and steering — set OKRs from the AI's grounded recommendations and then deliberately overrule one, on the record. Run the Growth Matrix estimator with its built-in devil's advocate. Install the weekly review where autonomy dials move on evidence.
- Scaling yourself — onboard your next human into the hybrid workflow, write the one-page position on what stays human forever, and then the capstone.
The capstone is a publication, not a certificate
The playbook does not end with a quiz. It ends with you publishing your own point of view on how you lead an AI-run marketing org — assembled from the charters, retros, and overrules you wrote along the way, shipped through the very content machine you just mastered.
That choice is deliberate, and it is the part I find genuinely new. Your published POV does two jobs at once. Externally, it is category-defining thought leadership that very few CMOs can write honestly today. Internally — and this is the syntropic part — it ingests back into your own T2D3 OS as PRIMARY signal. The machine that helped you become this kind of leader now grounds every future draft in your stated philosophy. The student's thesis becomes the teacher's textbook.
Why a template, not a feature
We could have written this as documentation. Nobody reads documentation into their calendar. A playbook template turns the operating model into scheduled, role-tagged, dependency-ordered work: twenty-nine tasks with durations, each one practiced on your live GTM, not on a sandbox. The intent-based-leadership week is five real days. The signal ritual books itself into your week. The overruled OKR is a real OKR.
It is also, frankly, our own thesis eating its own cooking. T2D3 OS is built on the belief that AI agents should lead with initiative while humans provide the judgment, the locks, and the glint. A product that believes that should be able to teach it — and the teaching should arrive the way everything else in the OS arrives: as a project with tasks, owned by you.
CMO Mastery: Leading T2D3 OS is live in the template library now, available on every plan. Add it to a project, and start with task one: walk the app and ask, at every screen — who is proposing here, and who decides?