Define readiness score formulas that go beyond gut feel

From gut feel to defensible readiness

If your team is still greenlighting interviews on vibes, you’re leaving outcomes to chance. The cost is real:

• Inconsistent referrals erode employer trust
• Slower time-to-offer because coaching lands too late
• Burned-out mentors spending hours on candidates who aren’t ready

A transparent readiness score changes the game. It aligns coaches and learners, focuses practice on the right gaps, and gives directors a weekly signal to forecast offers. With SageStepper’s Nova AI interviewer, progress radar, and action boards, you can turn raw practice into a confident go/no-go decision in under 10 minutes per learner.

Principle: Celebrate progress, not perfection—and make the path to “Go” unambiguous.

Define the score: a transparent formula you can publish

The most effective formulas blend four pillars. Keep it simple, explainable, and hard to game.

• Rubric mastery (40%): Median competency score across targeted rubrics (e.g., product sense, data depth, communication), 0–100.
• Complexity coverage (25%): How well the learner performs on increasing difficulty and diverse scenario types.
• Consistency over time (20%): Stability via 3-run rolling median and variance; streak length matters.
• Behavioral signal quality (15%): Concise structure, follow-up handling, time-to-first-answer, reflection quality.

Publish your weights and thresholds. Here’s a ready-to-use configuration:

# readiness.yml
version: 1
pillars:
  rubric_mastery:
    weight: 0.40
    metric: median_rubric_score      # 0-100 from Nova rubric
  complexity_coverage:
    weight: 0.25
    metric: complexity_index         # 0-100 across difficulty & scenario mix
  consistency:
    weight: 0.20
    metric: stability_score          # 0-100 from variance & streaks
  behavioral_quality:
    weight: 0.15
    metric: behavioral_index         # 0-100 (structure, follow-ups, timing)

bands:
  green: 
    min_score: 75
  amber:
    min_score: 65
  red:
    min_score: 0

# Early exit gates
minimums:
  sessions: 6          # require practice volume
  sprint_complete: true
  median_rubric_min: 70

ethics:
  exclude_features: [accent, name_pronunciation, time_of_day]
  subgroup_parity_range: [0.8, 1.25]

And an implementation you can adapt:

// readiness.ts
export type Metrics = {
  rubricMastery: number;        // 0-100
  complexityCoverage: number;   // 0-100
  consistency: number;          // 0-100
  behavioralQuality: number;    // 0-100
};

export type Context = {
  sessions: number;
  sprintComplete: boolean;
  medianRubric: number;
};

export function computeReadiness(m: Metrics, ctx: Context) {
  // Gates first
  const gatesOk = ctx.sessions >= 6 && ctx.sprintComplete && ctx.medianRubric >= 70;
  const score = 0.40*m.rubricMastery + 0.25*m.complexityCoverage + 0.20*m.consistency + 0.15*m.behavioralQuality;
  let band: 'GREEN'|'AMBER'|'RED' = score >= 75 ? 'GREEN' : score >= 65 ? 'AMBER' : 'RED';
  // If gates fail, force AMBER/RED for coach-first review
  if (!gatesOk) band = band === 'GREEN' ? 'AMBER' : band;
  const decision = band === 'GREEN' && gatesOk ? 'GO' : band === 'RED' ? 'NO-GO' : 'COACH-FIRST';
  return { score: Math.round(score), band, decision };
}

Why this works:

• It is explainable to candidates and employers
• It guards against “lucky spikes” via consistency checks
• It respects complexity—difficult interviews count more
• It never penalizes protected or irrelevant attributes

Collect the right signals in SageStepper (without extra spreadsheets)

SageStepper already captures the signals you need.

• Nova AI interviewer: per-competency rubrics, follow-up handling, time-to-first-answer, and calibrated difficulty tags
• Interview Sprint: 5–7 day ritual to accumulate 6–8 sessions quickly; track streaks automatically
• Progress radar: visualizes competency medians, complexity coverage, and stability over time
• Action boards: route Amber/Red learners to explicit next steps (flashcards, targeted drills, mentor mock)

Sample session record Nova produces under the hood:

{
  "sessionId": "sess_4821",
  "candidateId": "cand_90",
  "timestamp": "2025-10-18T15:04:23Z",
  "rubrics": {"communication": 78, "problem_solving": 72, "technical_depth": 68},
  "difficulty": "hard",
  "scenario_type": "system design",
  "time_to_first_answer_sec": 24,
  "followup_quality": 0.82,
  "structure_score": 0.76,
  "reflection_quality": 0.71,
  "nova_version": "v3.4"
}

From these, SageStepper computes the four pillar indices shown in the progress radar and pipes Amber/Red cases to action boards for coach review. See how it fits together on the workflow page: /workflow?utm_source=blog&utm_medium=cta&utm_campaign=readiness-score-formulas.

Fairness and clarity: ethics baked in

A score is only useful if it’s fair and understood.

• Be explicit about what is and isn’t measured. Publish your readiness.yml to students in your handbook.
• Exclude irrelevant attributes: accent, name_pronunciation, time_of_day, and device type.
• Show uncertainty: display error bars when sessions < 6 or variance is high.
• Run parity checks weekly: compare average scores and go/no-go rates across subgroups (e.g., program track, time zone, disability accommodations).

Example parity check you can run on your exports:

-- readiness_parity.sql
WITH latest AS (
  SELECT candidate_id, score, decision, subgroup
  FROM readiness_snapshots
  WHERE snapshot_date = CURRENT_DATE
)
SELECT 
  subgroup,
  AVG(score) AS avg_score,
  AVG(CASE WHEN decision='GO' THEN 1 ELSE 0 END) AS go_rate,
  AVG(CASE WHEN decision='NO-GO' THEN 1 ELSE 0 END) AS nogo_rate
FROM latest
GROUP BY subgroup;

Look for parity ratios between 0.8 and 1.25. If outside range, investigate feature drift, calibration, or practice access (e.g., time windows for mocks). Document decisions and adjustments in your readiness playbook.

Publish the "why" behind gate failures. Students accept hard feedback when the path forward is clear.

Decision bands, gates, and weekly rituals

Turn the formula into predictable decisions and coaching behaviors.

• Bands: Green (75+), Amber (65–74), Red (<65)
• Gates: ≥6 sessions, Interview Sprint complete, median rubric ≥70, last 3 medians within ±5 points
• Outcomes:
  • Green + Gates = GO (refer to employer or book live mock)
  • Amber = COACH-FIRST (assign targeted drills; recheck in 72 hours)
  • Red = NO-GO (reset goals; schedule mentor triage)

Run two low-lift rituals:

1. Interview Sprint (7 days):
   • Day 1: baseline Nova mock; align on role rubrics
   • Days 2–6: one mock/day + flashcards; focus on weakest rubric from progress radar
   • Day 7: Nova hard mock + live mentor mock (if Amber→Green trend)
2. Readiness standup (20 minutes, weekly):
   • Review progress radar for trend outliers
   • Triage Amber/Red on action boards; assign next steps
   • Log gate exceptions and parity checks

Decision logic you can paste into your SOP:

# decision_rules.yml
if: score >= 75 AND gates.met
then: GO
else_if: 65 <= score < 75
then: COACH-FIRST
else: NO-GO

recheck:
  amber: 72h
  red: 7d

exceptions:
  employer_deadline: allow GO with live mock pass >= 80
  disability_accommodation: extend Sprint to 10 days without penalty

Point coaches and students to the centralized workspace: /product?utm_source=blog&utm_medium=cta&utm_campaign=readiness-score-formulas.

30-day rollout plan (minimal meetings, maximal signal)

Week 1 — Define and publish

• Customize readiness.yml weights and gates
• Import rubrics for target roles in SageStepper; align with employers
• Train mentors on Nova feedback norms; set expectations for score transparency

Week 2 — Pilot with 20 learners

• Run a 7-day Interview Sprint cohort: /workflow?utm_source=blog&utm_medium=cta&utm_campaign=readiness-score-formulas
• Track streaks and medians; log coach time per learner
• Start parity checks; capture questions for FAQ

Week 3 — Calibrate and automate

• Compare Green/Amber/Red to live mock outcomes; adjust weights ±5% max
• Add error bars when sessions < 6; require gate completion before Green
• Pipe Amber/Red to action boards with auto-assignments (flashcards, drills)

Week 4 — Scale and report

• Expand to all tracks; run the readiness standup
• Report outcomes: readiness score lift, mentor hours saved, offer rate trend
• Publish iteration notes to build trust with learners and staff

What outcomes to expect (and how to explain them)

From recent cohorts using this model in SageStepper:

• +18 point readiness lift in 21 days (median 62→80) after two Interview Sprints
• 35% mentor hours saved by shifting early practice to Nova and using Amber triage
• +12% offer rate over prior term when Green gating preceded employer referrals

How to communicate to learners:

• “Your score moved from 64 (Red) to 71 (Amber) in five days. The blocker is consistency (variance ±12). Complete two medium mocks with the same framework; we’re one step from Green.”
• “Complexity coverage is 58. Let’s add two hard system design scenarios; Nova will simulate follow-up pressure so you can practice structured pivots.”

How to communicate to employers:

• “We refer candidates at 75+ with stable 3-run medians and hard-level performance. Attached: radar chart and decision log.”

If your context is unique (e.g., accessibility needs, limited internet windows), adapt gates without diluting fairness. SageStepper makes those exceptions visible in the decision log so you maintain consistency and compassion.

Templates you can reuse tomorrow

Readiness review agenda (20 minutes):

• Open progress radar; surface top 3 Amber trends
• Pull action board; assign 1–2 tasks per learner
• Check parity dashboard; log any out-of-range ratios
• Confirm next Interview Sprint start dates

Student-facing message template:

Subject: Your interview readiness plan for the next 7 days

Here’s where you stand today: Score {{score}} ({{band}})
What’s working: {{strengths}}
What to improve: {{gaps}}
Plan:
- Day 1–2: Nova mocks focused on {{weakest_rubric}}
- Day 3–5: Flashcards + targeted drills
- Day 6–7: Hard mock + reflection journal
We’ll recheck on {{date}}. Reply if you need accommodation—we’ll adjust gates without penalty.

Coach checklist for Amber learners:

• Confirm gates: sessions ≥6, Sprint complete
• Assign two mocks: one medium, one hard
• Add one 15-minute live drill on weakest rubric
• Schedule recheck in 72 hours; nudge via /login?utm_source=blog&utm_medium=cta&utm_campaign=readiness-score-formulas

Measure, learn, and iterate (without breaking trust)

Track these metrics in your dashboard:

• Readiness score lift (baseline → week 3)
• Offer rate and time-to-offer by band
• Mentor hours per learner (target 25–40% reduction)
• False Green/False Red vs. live mock outcomes (target <10%)
• Subgroup parity ratios and drift flags

Cadence:

• Weekly: parity + misclassification review; adjust features, not the goalpost
• Monthly: revisit weights ±5% with evidence; publish changes
• Quarterly: A/B test a gate (e.g., median rubric from 70→72); compare offer rate delta

SageStepper’s progress radar and exports make this turnkey. Explore the product: /product?utm_source=blog&utm_medium=cta&utm_campaign=readiness-score-formulas