Operator Notes

Field-tested thinking on the math behind the trade.

Long-form essays from the Eagle Rock team on governance, incentives, decision-making, and the math that drives them. Different register than our guides — opinionated, math-first, no CTA.

The Monthly Close Is a Tax on Truth

A close that lands on day 5 isn’t faster — it’s the same backward-looking artifact with a tighter deadline. Faster close still produces a record, not a report that changes anything.

Why Winning the Auction Usually Means You Overpaid

The auction mechanism selects for the highest bid, which comes from the bidder with the most wrong estimate. Winner’s curse has been known for sixty years; nobody’s fixed it.

Treat Your Next Big Move Like Three Small Ones

Committing to the full version of a big move before the small version is proven is the most expensive mistake in ambitious undertakings. Real options, kill criteria, phased commitment.

AI Operator Isn’t a Job — It’s Two Jobs

Two jobs share the title “AI operator.” One builds AI systems; the other uses AI in their actual job. They share no skills, no risks, no vocabulary. Hiring one role to do both is the most expensive category error in modern org charts.

5-Year CEO Comp Windows: What Boards Are Actually Buying

Same per-year comp, wider range of outcomes, no extra compensation for the longer uncertainty horizon. Boards call it long-termism. The math says it’s a variance trade.

What You’d Bet vs What You’d Say: Why Confidence Breaks

If you’d only bet 50% on a deal you said was 90%, you don’t have a confidence problem — you have a feedback-loop problem. Calibration is a skill, not a trait.

The Org Chart Is Fiction: Here’s Who Actually Decides

The chart on the wall describes reporting lines, not decisions. In any company of real size, the two answers diverge — and the cost is 10–20% of strategic decisions routed through the wrong room.

Decisions as a Portfolio: Stop Punishing Individual Calls

A 60% hit rate across 10 strategic decisions isn’t failure — it’s a good year. Evaluating each decision in isolation is the cognitive error that makes leaders punish winners.

Cash Reserve for the Events You Can Already See

Your biggest customer’s contract is up in 18 months. Your lease renews at a known number. Your best engineer has been interviewing. The events aren’t surprises — the cash just wasn’t reserved against them.

The Cheapest Edge Is Being Unreadable

Visibility to the market is visibility to your competitors. Being unreadable is the cheapest strategic edge — and most companies trade it away for a brand campaign.

Bad at Cutting Losses, Bad at Walking Away

The sunk-cost fallacy is the one you’ve heard of. The holding fallacy — continuing because time has already been invested — is the one you haven’t. They’re both reasons you don’t walk away when you should.

Your Financial Model Is Probably Lying to You

Revenue projections, attribution reports, AI benchmark tables — they all look like measurements. What they actually are is a small number of guesses dressed in mathematics.

How Judgment Actually Gets Built

Judgment isn’t an innate trait. It’s a learnable practice: small bets, visible outcomes, written records. Most people skip the practice because the work feels like failure.

Why AI Failures Are Permission Problems, Not Model Problems

Most AI failures aren’t capability problems. They’re permission problems — the system can do the thing, but nobody decided what it should be allowed to do without a human in the loop.