course module
The 4D Playbook
The four collaboration quadrants and the specific rework each stuck project needs.
Sandbox
course module
The four collaboration quadrants and the specific rework each stuck project needs.
Born in the Optimus Mastermind. This kit came straight out of the July 8 session — the SaaS-renewal thread and the collaboration rethink for the agent era. The full call — and the room where these get built live, alongside other founder-architects — lives inside the Optimus Mastermind. Apply to join at buildwithoptimus.com →
Member edition. Four quadrants. Three rework patterns. One template you fill in per project.
This playbook picks up where the free diagnostic leaves off. The diagnostic tells you which quadrant each of your projects lives in. This tells you how to move them — the specific meeting to cancel, the specific approval to cut, the specific written waypoint to introduce, and the specific way an owner takes back discernment and diligence from the committee.
It's written for a $5–50M founder-architect who has already given agents to their team and can't figure out why velocity hasn't moved. The answer is almost never the agents. It's the collaboration model layered on top of them.
Almost every $5–50M business has the same shape:
The rework, quadrant by quadrant, is below.
The project has a named owner. Ownership is real. But the collaboration model is a weekly (or twice-weekly) meeting where the team weighs in on progress. The meeting is on the calendar because the team asked for it, or because the owner wanted air cover, or because it's always been there.
The problem: the meeting has become the pacing mechanism. The owner doesn't ship between meetings because they know the meeting is coming. The team doesn't review between meetings because they know the meeting is coming. Cadence and quality are now hostage to a calendar hold.
Move 1 — cancel the recurring meeting. Kill it on the calendar. Send a Slack/Signal note: "Recurring standup is off. Waypoint reviews on demand."
Move 2 — install written waypoints. The owner defines three-to-five waypoints for the project (e.g., "concept approved," "v1 draft," "v2 tested," "shipped"). At each waypoint, the owner writes a two-paragraph summary — what shipped, what's next, what feedback they need — and drops it in the project channel. Reviewers respond on their own time within 48 hours. No meeting required.
Move 3 — reserve an on-demand review slot. The owner gets a standing "if needed" 30-minute slot per week that the team keeps open. If a waypoint sparks a decision that genuinely needs synchronous discussion, the owner uses the slot. If not, the slot expires unused. Most weeks, it will.
If a waypoint attracts unresolved disagreement in writing (three or more comments contest the same decision), the owner escalates by scheduling the on-demand slot within 24 hours. Escalation is the owner's tool, not the reviewers'.
A project that shipped every two-to-four weeks under a weekly meeting typically ships every three-to-seven days under written waypoints. Ownership becomes real; the meeting was the bottleneck all along.
Every step of the project routes through someone else's inbox for sign-off. Legal reviews the copy. Finance reviews the spend. IT reviews the tool choice. Compliance reviews the process. The owner's role has been reduced to project-managing a queue of external gatekeepers.
The problem: the critical path is now the slowest inbox. And most of those sign-offs are not genuinely required — they're precautionary sign-offs that got layered in during a prior incident and never got removed.
Move 1 — audit the sign-off list. For each approval, ask: what specifically would go wrong if we skipped this? If the answer is "nothing specific, but we always do it," cut it. If the answer is a specific incident from three years ago that no longer applies, cut it. If the answer is a specific real risk today, keep it — but move it.
Move 2 — collapse per-step approvals into one deliverable-review at the end. The owner runs the full project autonomously, then presents the deliverable to the two-to-three named reviewers whose sign-off is genuinely required. Everyone else is FYI, not gatekeepers. One review, at the end, against the finished thing — not four reviews scattered through a half-finished thing.
Move 3 — write the escalation contract. The owner drafts a short "when I will escalate" contract at kickoff and hands it to the reviewers. Something like: "I will escalate to Legal if the copy names a competitor by trademark or makes a substantiated-claim assertion. Otherwise, I will not escalate mid-project." This gives reviewers explicit permission not to be in the middle.
Some approvals genuinely have to be inline — the ones with regulatory or contractual weight (SOX sign-offs, HIPAA reviews, etc.). Leave those alone. Cut the precautionary ones.
A project that took eight weeks under per-step approvals typically ships in two-to-three weeks under one-deliverable-review. The owner gets ownership back; the reviewers get their inbox back.
No named owner. Everyone touches the project a little; nobody feels responsible for the whole. Meetings are the work — the project moves forward one meeting at a time because that's when the group is together.
The problem: this is the highest-leverage rework and the hardest one, because the committee itself is the problem, and the committee will resist being disbanded.
Move 1 — pick an owner. One human. Named. Not a co-lead. Not a team. One. If nobody on the committee wants the job, that is the first useful signal the diagnostic has produced — this project may not actually be a priority, and the honest move is to shelve it rather than committee it forward.
Move 2 — give the owner a bench of agents. Marketing project → marketing agents. Legal project → legal agents. Ops project → ops agents. The owner assembles the bench that matches the project scope — inside Optimus, or with your own agent stack.
Move 3 — dissolve the committee. Send a note: "I am owning this end to end. I will drop a written waypoint in this channel when I hit each of these five markers. Please respond in writing within 48 hours if you have feedback. There will not be a recurring meeting for this project." Then hold to it. The first two weeks, half the committee will try to re-summon the meeting. Don't.
The rewrite works because the owner earns trust in real time by shipping. Waypoint one lands and is well-executed → committee members exhale. Waypoint two lands → they stop asking. Waypoint three lands → nobody misses the meeting. Trust follows shipping, not the other way around.
A project that has been in committee for six months and hasn't moved typically ships its first waypoint within seven-to-fourteen days under a single owner. The largest cycle-time delta in the playbook by orders of magnitude.
The 4D Contract is a one-page template the owner completes at project kickoff. It's the companion to the free diagnostic. It follows the 4D framework — Delegation, Description, Discernment, Diligence — turned into a fillable contract.
Field 1 — Delegation. For each step of this project, is a human doing it or an agent? The owner lists the project's five-to-ten steps and marks each Human or Agent. This is the first useful lens because it forces the owner to see what actually parallelizes.
Field 2 — Description. What is the definition of done, in one paragraph? Not a brief. A definition. If v1 doesn't match this description, v1 isn't done. If v1 matches this description and the committee wants more, the committee is scope-creeping — the owner points at the description and says "we agreed this was done."
Field 3 — Discernment. Who judges whether the output hits the bar? The owner names themselves. If they don't feel qualified to judge, either they're the wrong owner or they need a specific-named reviewer whose taste they defer to (not a committee, one person). Discernment lives with one human always.
Field 4 — Diligence. What's the checklist for accuracy, bias, edge cases — and security and data protection — the owner runs before shipping? Three-to-seven bullets, specific to the project. In the agent era this field is where security and data-protection guardrails live, and they are non-negotiable: which data and systems the agents were allowed to touch (least privilege — nothing more than the job needs), that no customer PII, secrets, or credentials leaked into prompts, logs, or outputs, and that nothing an agent produced opens a prompt-injection or data-exfiltration path into production. Handle data per your own privacy and retention policy (GDPR / CCPA / SOC 2 as they apply to you). The owner runs the checklist — especially the security items — and cannot delegate it to the agents or the committee. Diligence is the owner's non-negotiable.
The 4D Contract lives at the top of the project channel. When anything goes sideways, the team reads the contract, not the meeting notes.
Most founder-architects run the rework in this order:
Total elapsed time: about a quarter. Total cycle-time reduction for a typical $5–50M founder-architect running all three: 40–70 percent across the project portfolio, without adding headcount.
Every rework in this playbook is the same shape:
Follow the pattern and cycle time collapses without shrinking your team. Skip it, and adding agents to a committee just gives the committee more to argue about.
Once you've moved every active project into End-to-End Owner, run the diagnostic again on the pattern itself — which owners are shipping fastest, which agent skills keep coming up, which waypoints turn out to matter.
Not in Optimus yet? Apply to join the Optimus Mastermind → — it's where these reworks get built live, alongside other founder-architects.
Already a member? Bring the diagnostic output for your three most-stuck projects to the weekly help call on Wednesdays. We'll pick one, fill in the 4D Contract live, and give you the sequence to run against the other two.