06-reference

commoncog mckinsey management consulting creation

2026-07-13·reference·source: Commoncog·by Cedric Chin

How to Build a $3.3 Billion Dollar Consulting Company ... From Scratch

Why this is in the vault

McKinsey is the canonical example of consulting-as-expertise-codification at scale. Cedric Chin's two new cases — "The Origins of McKinsey" (free) and "Scaling McKinsey: Inventing Management Consulting" (members-only) — trace how James Oscar McKinsey and Marvin Bower turned a single accounting insight into a global methodology machine. The question Chin poses — does McKinsey have a durable moat, and what is its source — is the same question RDCO is implicitly asking about the harness-engineering thesis.

The core argument

The McKinsey arc runs in two phases. In the founding phase (1926), James Oscar McKinsey invented "budgeting" as a management discipline, transforming accounting from bookkeeping into a forward-looking control tool. That single reframe gave McKinsey its wedge: not accountants, but advisors with a methodology. The second phase belongs to Marvin Bower, McKinsey's successor, who deliberately positioned management consulting as a profession — with rigor, standards, and a replicable problem-solving system. Bower's strategy was deliberate; the World War II demand surge was opportunistic. Chin's core analytical question: has McKinsey's moat held across a century, and what was it actually made of?

Curation section

Cases announced this issue:

External links:

Member forum highlights: Moat identification difficulty on "invisible companies" (cornered resource hiding in plain sight); investing expertise + NDM knowledge management as a decision-tracking system; Swiss watch capital allocation comparing Hayek vs. Richemont's acquisition strategy; writing expertise improvement program in progress.

Mapping against Ray Data Co

Marvin Bower's core move — encoding consulting judgment into a firm's way of working — is the direct historical predecessor to what RDCO's harness-engineering thesis is attempting. McKinsey's durable asset wasn't individual genius; it was structured problem-solving methodology, client development protocols, and people-development systems that made expertise legible, reproducible, and transferable at scale. The CLAUDE.md + skills + vault + SOPs stack is RDCO's equivalent: Ray's tacit operational judgment packaged into a methodology the Claude-as-COO agent can execute without reconstructing from scratch each session.

The sharpest tension Chin surfaces is the moat durability question. McKinsey's methodology advantage eroded as competitors replicated the form (slide decks, frameworks, partner model) without the underlying rigor. RDCO's analogous risk: the harness-engineering methodology layer gets commoditized as frontier models improve and everyone ships CLAUDE.md clones. The McKinsey story suggests the moat in consulting came from people development and selection standards as much as from methodology — the process for inducting and calibrating practitioners. For RDCO this raises a specific open question: what is the analog to McKinsey's analyst/associate training pipeline in an AI-native operation?

Secondary connection: Chin's note that McKinsey's initial wedge was a reframe (budgeting as management tool, not bookkeeping) maps to Ray's phData DSA positioning — the wedge is framing AI deployment as a methodology discipline (targeting systems, outcome procurement) rather than a tooling discipline.

Related