11 months · DTC · ~mid 8 figure / yr
Installed an operating system that holds while the founder takes his first sabbatical in a decade.
The problem
The CEO was the company.
Ten years solo-founding a premium DTC brand to mid-8 figures had left Mike Bell carrying every operational decision in his head. Execution was force-of-will. Trade-offs were implicit. Mistakes got paid for in cash and product runs. A regulatory event or supplier failure would have pulled him into a foxhole he couldn't climb out of.
He needed to take a sabbatical. The company couldn't survive him taking one.
"Our biggest problem is execution." — Mike Bell
The solution
Get Mike out of the operational foxhole. Build the system that runs without him.
Eleven months, sequenced by department dependency. Each install unlocked the next:
- Marketing. Built the go-to-market systems — the cycle from insight to launch, channel by channel. This surfaced the deeper issue: every department was running its own filing system. Information couldn't travel.
- Single operating system. Two weeks transferring the whole company onto Notion. 83 people onboarded. 8.96/10 onboarding experience. 9.15/10 on "I feel excited to use Notion." The robot can't work on what it can't see — and neither can the next operator who walks in.
- The Stage Gate. Mapped the full company lifecycle from product inception to delivery. 210 steps. 9 departmental owners. Every step assigned, every handoff documented, every dependency surfaced. Stage 0 made non-negotiable. The Stage Gate made project management legible for the first time — every project visible, every owner explicit, every blocker surfaced.
- Project management department. Built the function from scratch. Air Traffic Control as the operating rhythm — weekly cadence, cross-functional visibility on every project in flight. Hired the PMs to run it. Now scoping against reference-class data — real distributions of how long this kind of project actually takes, not founder intuition.
- Acting Head of R&D. Stepped into a hole that opened once the operational backbone stabilised. Transitioned the innovation team into a real R&D function. Built the 3-year pipeline with Mike — grounded in reference-class data and real cost estimates. Variance of risk attached to every project based on prior reference cases.
- Data mapping. Once decisions were running through real reference classes, the bottleneck moved to the source data feeding them. Mapped every system Manukora ran on — Cin7, Shopify, Amazon, Meta, GA, Klaviyo, the spreadsheets in between. Identified where data was clean, where it was dirty, and where the pipes between systems were leaking. Stock on hand and stock in market unified.
- Supply chain adoption. With data flow architecture in place, the final wave was getting the factory floor inputting cleanly at point-of-event rather than retroactively. Rigorous Notion adoption across the supply chain team. Logging standards installed.
Handoff was Feb 2026. ATC ritual, S&OP cadence, NPD lifecycle, research discipline — all still running 2.5+ months out.
The exit was the deliverable. Talk to Mike.
What I didn't do
I didn't fully install the Decision Engine at Manukora — and I want to be clear about why.
My mission was to get Mike out. Mike's now out. The work that compounds — Stage Gate, Notion adoption, ATC, S&OP, the customer research engine — was the right substrate to install given what the tooling could do at the time. Everything I built put Manukora in the position to receive the brain the moment it became viable. Most companies aren't there. Manukora is.
The engine became viable in April 2026.
The departmental LLMs, the always-on cognitive layer, the cross-functional correlations surfaced in real time — that capability has only matured in the past few months. After Manukora, I came home and pushed it to the limit on the only testbed I had unconditional access to: myself.
Built my own brain. 69 hours over 23 days. It's a direct reflection of how I think, what I track, and what I want surfaced. The architecture is the same one I'd install in your business — training corpus, decision-making process, departmental dashboards, brain on top — running daily across my work, finances, learning, and accountability.
This isn't a case study because the timeline doesn't allow one yet. The capability is six weeks old. What I'm offering is a system I've stress-tested on the highest-stakes testbed I have, ready to install in a business the moment the founder is.
→ See Full Overview Of Completed Work
The proof?
Today (May 2026): Mike is working 1–2 hours a week. The system holds.
That is the case study. The work speaks for itself.
