A title is only useful if someone can hire against it. The case for the Predictionist as a role has been made elsewhere on this site, in the argument that the seat already exists and in the read on which roles cross over into it. What has not been written is the job description. This is it, in the shape a CEO or a head of people would actually need: what the person owns, by phase, who they answer to, what they are measured on, and what the first ninety days look like. Compensation is deliberately left out. That depends on the company, and anyone who gives you a number without knowing the company is guessing.
The one-line mandate
The Predictionist runs Woodring’s Loopon behalf of the leadership team. Their product is not a forecast and not a deck. It is the organization’s re-priced posture, refreshed at the cadence the operating environment requires, with every commitment sized to the cost of being wrong. The controller is accountable for the books. General counsel is accountable for the contracts. The Predictionist is accountable for whether the company is reading the future well enough to commit capital with its eyes open.
Responsibilities by phase
Scan. Own the signal layer. That means a curated set of external series, internal operating indicators, and human signal from sales, supply, and the field, each with a named source and a refresh cadence. Weekly, produce a half-page read of what moved. The Predictionist decides what gets watched and, just as important, what gets dropped. A signal layer that only grows is a signal layer nobody reads. They also own the AI/Human Split for this phase: the machine does the volume work of ingestion and pattern surfacing, and a named person decides what it means.
Story. Maintain the active scenario set. Three to five named worlds on real cleavages, each with an indicator set and a short written narrative the executive team can rehearse against. Monthly, the Predictionist reads the scenarios against the signal layer and proposes which are gaining and which are losing support. They write the scenarios. They do not decide alone what the company does about them.
Stake. Prepare every material commitment for sizing. Before a capital decision, a hiring wave, a multi-year contract, or a market entry reaches the executive sponsor, the Predictionist has already placed it on the Stake Sizing Matrix: which scenario it is sized for, how reversible it is, how much calibrated conviction the team actually has. They run the blind-vote conviction read. They do not own the budget. The executive sponsor commits; the Predictionist makes sure the commitment is sized to the risk shape and not to the rank of whoever is asking.
Steer. Chair the monthly Convergence Review. Every active scenario lands in one of three columns, converging, holding, or collapsing, and every stake attached to a scenario that moved gets re-examined in the room. The Predictionist owns the agenda, the three column board, and the written list of stake adjustments with a named owner for each. They also own the follow-through: did the adjustment actually happen, or did it get absorbed back into the old plan.
Who they report to
The CEO, with a working line into the CFO. Not the other way around. A Predictionist who reports into finance becomes an FP&A function with better vocabulary, because finance owns the numbers and the Loop is about the story and the stakes underneath the numbers. A Predictionist who reports into strategy becomes a project manager for the offsite. The role needs to sit where the commitments get made, and in most companies that is the CEO’s staff. The relationship with the CFO is the one to get right, and the CEO relationship is the one that decides whether the role survives its first budget cycle.
Direct reports are few. One or two analysts who run the signal plumbing, and a rotating seat from each operating unit that owns a scenario. The Predictionist is not building an empire. The leverage comes from running a cadence the executive team keeps, not from headcount.
What they are measured on
Not forecast accuracy. A forecaster is graded on hitting the number. A Predictionist is graded on whether the organization’s commitments were sized correctly against the range, and on how quickly the organization moved when the range changed. Four measures do the work.
- Decision latency. The gap between a signal first appearing in the layer and the commitment it should have changed actually changing. This is the number the Cycle Auditestablishes on day one, and the Predictionist’s job is to shrink it every quarter.
- Calibration. When the team said it had high conviction, was it right at roughly the rate high conviction implies? Track it per scenario. A team that is confident and wrong is the most expensive team in the building.
- Stakes moved. How many commitments were grown, reduced, or killed at Convergence Reviews because a scenario moved columns. Zero for three months in a row means the meeting has become a status update.
- Surprise count. Material events that hit the company with no scenario in the active set that covered them. The goal is not zero. The goal is that each one produces a new cleavage, not a new excuse.
What the role is not
It is not the chief data officer. The Predictionist consumes the data platform; they do not build it. It is not the head of FP&A. Rolling forecasts and variance analysis stay in finance, and the Predictionist uses their output as one input into Steer. It is not the futurist on the speaking circuit. Everything the role produces is tied to a commitment somebody in the company has to make, which is the difference between foresight with skin in the game and narration.
The first ninety days
The first revolution of the Loop is the onboarding plan, and the ninety-day installation is already written in detail. In short: days one through thirty build the signal layer against one bounded operating question, not the future of the industry. Days thirty-one through sixty produce three named scenarios with indicator sets. Days sixty-one through ninety end in a real stake decision with the executive sponsor in the room and a standing Steer cadence that survives the quarter.
Two things need to be true before the person starts. The CEO has to have committed to attending the Convergence Review, every month, in person. And a decision with real money attached has to be inside the ninety-day window, because a Predictionist who finishes the first quarter with a beautiful scenario set and no commitment has proven the role is a planning exercise, which is exactly what it must not be.
Before you post it
Run the Maturity Read with your executive team first. It takes six minutes and tells you which phase of the Loop is absent, which is episodic, and which is already functional. Hire for the gap. A company with a strong Scan and no Stake discipline needs a different Predictionist than a company with annual scenarios and no signal layer. The full role, including how it fits into the team structure in Part III, is in the book.