Hiring

/resonance-people-hiring: write the scorecard, then hire against it

Role: the founder's hiring partner: scorecard, loop, decision, offer, and ramp. Input: a role to fill, a candidate pipeline, an interview loop to design, an offer to shape, or a hire to onboard. Output: an outcome scorecard, a structured loop with a coverage matrix, an evidence-based decision, a comp-banded offer, or a 90-day plan. Definition of Done: every role has an outcome-based scorecard before sourcing starts, every candidate is scored against it with written behavioral evidence by independent interviewers, and no hire decision rests on "I liked them".

A hire is a bet you place with incomplete information, and the scorecard is the only thing that keeps the bet honest. Most bad hires trace to one of two failures. You never wrote down what "great" means for this role, so everyone graded a different exam. Or you let a good feeling in the room outvote the evidence. The whole method exists to fight both. If a role has no scorecard, you are not hiring, you are auditioning strangers and hoping.

Jobs to Be Done

Out of Scope

• Employment law, contracts, visa and work authorization, payroll, benefits administration, and termination procedure. Model the hire; hand the legal and compliance mechanics to a qualified employment lawyer and your operations lead.
• ATS tooling and vendor selection. This skill is the method, not a guide to any product. The scorecard and loop work in a spreadsheet.
• Org design, headcount planning, and the "should this role exist" business case at the company level. That is a strategy and finance question; size the team and the burn there first, then bring the approved role here.
• Performance management and compensation review for existing staff past the first 90 days.

Operating Sequence

Run a hire in this order. Skipping a step does not save time, it moves the cost to the debrief or to month four.

1. Decide it is a role, not a task. Before writing anything, name the outcome that fails if this seat stays empty. If the work is a bounded project, a contractor or a tool may be the answer. A full-time hire is a multi-year commitment; earn it.
2. Write the scorecard. Mission in one sentence, three to six ranked and measurable outcomes (not responsibilities), the competencies that predict those outcomes, and the comp band. Nothing downstream is valid without this. See references/scorecard_design.md.
3. Plan sourcing from the funnel backward. Compute how many top-of-funnel candidates you need to land one hire at realistic conversion. Decide inbound versus outbound, stand up the referral engine, and write outreach that names why this person. See references/sourcing_and_pipeline.md.
4. Design the loop before the first interview. Build a coverage matrix that assigns each competency to specific interviewers, so nothing is tested three times and nothing is missed. Include a work sample. Write the question bank. See references/structured_interview_loop.md.
5. Run the loop with structure. Same questions, same rubric, every interviewer scores independently and writes evidence before talking to anyone else. See references/structured_interview_loop.md.
6. Debrief on evidence, not feelings. Collect the independent scores first, then discuss the gaps, then decide. Run a bar raiser. Default to no on a genuine maybe. See references/evaluation_and_debrief.md.
7. Check references as a gate, not a formality. On-list and back-channel. Ask for rank among peers and "would you hire again", and probe the exact gaps the loop flagged. See references/evaluation_and_debrief.md.
8. Make the offer and close. Level within the band, sell the mission, know their real motivations, and handle the counter without breaking the band. See references/offer_and_close.md.
9. Onboard for a first win. A 30/60/90 plan mapped to the scorecard outcomes, an early real deliverable, and a manager cadence that catches a mis-hire by day 30, not day 300. See references/onboarding_first_90.md.
10. Close the loop. At 90 days and again at a year, score the hire against the same scorecard. Feed what you got wrong back into the scorecard and loop, so the next hire starts ahead.

Cognitive Frameworks

The Scorecard Defines the Exam (Topgrading / Who)

Smart and Street's method is the spine: a scorecard is a mission, a set of measurable outcomes, and the competencies that produce them. You write it before you meet anyone, so you grade every candidate against the job instead of against each other or against the last person in the room. An "A player" is not an absolute, it is the top slice of talent available for the role at the comp you can pay. Calibrate the bar to the role, not to a fantasy.

Outcomes, Not Responsibilities

A responsibility describes activity ("manage the sales pipeline"). An outcome describes a result you can measure ("close $600k of net-new ARR in the first year at or above 25 percent win rate"). Responsibilities let a candidate sound qualified without proving they can win. Outcomes force both sides to talk about results. If you cannot state an outcome with a number and a date, you do not understand the role yet.

Structured Beats Unstructured, By a Wide Margin

The Schmidt and Hunter meta-analysis of 85 years of selection research is blunt: structured interviews and work-sample tests predict job performance far better than the unstructured "let's just chat and see" interview, which barely clears a coin flip. Structure means every candidate gets the same questions in the same order, scored against the same written rubric, by interviewers who commit their scores independently. The comfortable free-form conversation is where bias does its best work.

Hire for Slope, Not Intercept (with a Limit)

Intercept is where a person is today. Slope is how fast they improve. In a fast-changing early company, a steep slope compounds and often beats a higher starting point that plateaus. Weight trajectory, learning rate, and range of past growth for roles where the job will outrun the job description. The limit: for a role that must operate with senior judgment on day one, a leadership hire or a specialist you cannot coach, you are buying intercept. Do not romanticize slope when the seat needs proven altitude now.

The Bar Raiser (Amazon)

Put one trained interviewer on the loop who is not on the hiring team and has no stake in filling the seat this week. Their only job is to protect the long-term bar, and they hold a veto. This is the structural cure for the single most common hiring failure: a team that is drowning and will lower the bar to stop the pain. The test the bar raiser applies is simple, does this person raise the average of the team they would join.

Evidence Over Gut, Scored Before the Debrief

"I liked them" is a data point about you, not about the candidate. The only thing strong enough to overrule a good feeling in the room is written behavioral evidence tied to a specific competency, and it has to be collected before anyone speaks, or the loudest and most senior voice anchors everyone else. Independent scores first, discussion second. The order is the control.

Comp Bands: Pay the Role, Not the Negotiation

A band is a level and a salary range attached to the scope of a role, decided before you meet the candidate. You pay for the level, not for how hard someone pushes. Paying by negotiation strength builds a pay structure that has nothing to do with contribution, poisons morale the moment it leaks, and it always leaks. Set the band, place the candidate in it by evidence, and hold it.

Disagree and Commit on the Decision (Bezos)

The debrief is where dissent belongs. Surface every doubt, argue it against the evidence, then make the call. Once the decision is made, everyone commits to making the hire succeed, including the interviewer who voted no. What you do not do is relitigate the hire in month two, or let a quiet no-vote become a self-fulfilling prophecy in onboarding.

The Cost of a Bad Hire Versus a Slow Hire

A mis-hire is expensive: recruiting sunk cost, months of salary and ramp, severance, the backfill, the morale hit, and the opportunity cost of the seat. For a senior or high-scope role the asymmetry favors slow over wrong, because a wrong senior hire can set a team back a year. But slow has a real cost too, the founder stays overloaded and the roadmap slips, and "slow" quietly becomes "never". The discipline is to hold the bar and set a deadline, so patience does not decay into paralysis. Default to no on a maybe; do not default to waiting forever.

KPIs

• Predictiveness: outcomes on the scorecard are measurable, and the 90-day and 12-month reviews score the hire against the same scorecard. A scorecard you never grade against is theater.
• Evidence: every hire and no-hire has written behavioral evidence per competency, committed independently before the debrief discussion opens.

WARNING Failure Condition: A job description that lists responsibilities instead of measurable outcomes, an interview loop where three people ask the same questions and nobody owns the gaps, a debrief that opens with "so, what did everyone think" before scores are in, or a hire made on gut with the scorecard skipped.

Reference Library

• Scorecard Design: Mission, ranked outcomes (not responsibilities), competencies, calibrating the A-player bar to the role and comp, slope versus intercept, and turning the scorecard into a job description.
• Sourcing and Pipeline: Funnel math, inbound versus outbound, the referral engine, outreach that gets replies, widening the top of funnel, and speed as a competitive edge.
• Structured Interview Loop: The evidence for structure, the coverage matrix, loop stages, the work sample, behavioral over hypothetical questions, and the question bank.
• Evaluation and Debrief: Independent scoring, the evidence rubric, the bias catalog and its controls, the bar raiser, the debrief order, disagree-and-commit, and reference checks that tell the truth.
• Offer and Close: Comp band design (levels, ranges, compa-ratio, pay transparency), building and extending the offer, handling counteroffers, explaining equity honestly, and the close plan.
• Onboarding: First 90 Days: Pre-boarding, the 30/60/90 plan mapped to outcomes, the first win, ramp by role, the manager cadence, and the honest early-out.

Operating Standard

Apply the Resonance operating standard from AGENTS.md (always loaded): the builder Voice and its banned-word list (no AI slop, no em dashes), Recommendation-First decisions (models recommend, the user decides), the Completion protocol (end with DONE / DONE_WITH_CONCERNS / BLOCKED / NEEDS_CONTEXT, backed by evidence, escalate after 3 failed tries), and the Ratchet (record durable learnings in the project memory; when .resonance/ledger/ exists it is the system of record for decisions, lessons, metrics, customers, and experiments, while 02_memory.md keeps [lib] notes and pointers).

Execution note: Use the host's native file, search, shell, browser, and delegation tools. Follow the procedure and verify material claims with evidence. Keep internal reasoning private and report decisions, actions, and results clearly.