06-reference

commoncog data driven will not skill

Tue Apr 14 2026 20:00:00 GMT-0400 (Eastern Daylight Time) ·reference ·source: Commoncog ·by Cedric Chin

“A Big Part of Being Data Driven is Will Not Skill” — @CedricChin

Why this is in the vault

Chin uses the Will Guidara / Hani Ichkan case from Restaurant Associates to make the argument that data-driven operation is gated by commitment to a cadence, not by tooling or statistical sophistication. For RDCO’s Playbook + Coaching posture, this is directly load-bearing: the deliverable is a weekly sensemaking habit, not a dashboard.

The core argument (paraphrased)

Chin tells the story of Will Guidara’s pre-Eleven Madison Park detour. After two years at Danny Meyer’s USHG — a paradigmatically “restaurant-smart” company — Guidara turned down an AGM promotion on his father’s advice and went to Restaurant Associates (RA), “the basement at one of the least” glamorous jobs in New York. His father’s distinction: restaurant-smart companies have autonomy and warmth but lack back-end systems; corporate-smart companies have the systems but suppress hospitality. Guidara needed the other half of his education.

At RA, Guidara worked mornings in purchasing under Ken Jaskot (counting oysters, receiving deliveries, calculating COGS) and afternoons in accounting under financial controller Hani Ichkan. The pivotal episode: Ichkan spotted food costs up two months running at one restaurant, cross-referenced a lobster-sales report, confirmed a supply-driven price spike with purchasing, got the chef to swap in a scallop dish, then pulled lobster from every RA menu in the group. Pen and paper. Leather-bound ledgers. “Red-ink meant actual ink.”

Chin’s punchline: “Ichkan … was more data-driven than the vast majority of tech business operators I know.” No dashboards, no pipelines, no real-time anything. What Ichkan had was (1) the authority to act on his sensemaking and (2) — more importantly — a commitment to look at the numbers every week.

Chin connects this directly to Amazon’s Weekly Business Review: strip away the mechanism design and the WBR’s simplest principle is “look at your numbers every week.” Monthly is too little. Daily is too much initially. Weekly is just right. Everyone can do this — copy-paste into a spreadsheet on Mondays, then look. “This is a matter of will, not a matter of skill.”

The ask, therefore, is not tooling or training. It is the decision to commit to a cadence of sensemaking — and the organizational permission to act on what you see.

Mapping against Ray Data Co

Mapping strength: strong. This piece names the central failure mode of the consulting business.

1. The deliverable is a cadence, not a dashboard. Our 2026-04-14-levie-agent-deployer-role-jd Posture 2 (Playbook + Coaching) has to be structured around installing a weekly review ritual the client runs without us. If the coaching engagement ends with a shiny MAC dashboard but no recurring Monday meeting, we’ve shipped the skill artifact and skipped the will artifact. The Ichkan test: can they run the review on paper if the tooling breaks?

2. MAC is Ichkan’s ledger, not a new category. The ../01-projects/data-quality-framework/testing-matrix-template MAC framework — the 3×6 scope/basis matrix with Stop/Pause/Go tiers — is structurally just the weekly review for AI-era data models. Chin’s case argues we should stop selling MAC as a sophisticated methodology and start selling it as “here is your Monday habit.” The simpler the framing, the more likely it sticks.

3. “Authority to act” is the under-specified half of state-ownership. ../04-tooling/rdco-state-ownership-architecture argues the client owns the vault + skills + data. Chin’s case adds the missing piece: Ichkan could act because RA’s structure gave the financial controller power to override a chef. Ownership of artifacts without organizational permission to act produces dead vaults. Engagement scoping needs to name the decision-rights change up front — who gets to call a Stop signal, and whose override does that supersede?

4. The agent-deployer role is Ichkan. Per 2026-04-14-levie-agent-deployer-role-jd, the agent deployer instruments workflows and manages evals. The Ichkan parallel is exact: cross-functional, ledger-keeping, with a standing authority to pull the equivalent of lobster off the menu when the weekly numbers say so. When hiring or scoping this role for a client, the question is not “do they know statistics” but “will the CEO back their Stop call at 9am Monday.”

5. Against the phData/MG posture. The pure-delivery consulting model (build the platform, hand it over, leave) is the opposite of installing an Ichkan. The phData/MG outcome is a restaurant-smart company with a bolted-on corporate-smart artifact that nobody runs. Chin’s case vindicates the Playbook + Coaching posture: we don’t leave until the client’s own Ichkan is running the Monday review with us silent in the room.

One thing this article adds over the first-principles piece. 2026-04-15-commoncog-becoming-data-driven-first-principles gave us the epistemology (knowledge over truth, variation over vibes). This piece gives us the mechanism of change: commit to weekly, act on what you see, do it with whatever tools you have. If the first-principles piece is the theory, this is the installation instruction. The consulting product is the weekly cadence; everything else is scaffolding.