Back to the kit
From the talk

Scorecard Design: The 4-D Model

Rate each initiative from 1 (Low/Poor) to 5 (High/Excellent).

1. The AI Fit Test (Complexity vs. Determinism)

  • The Question: Does this require probabilistic reasoning, or could an if-else statement solve it?
  • Scoring:
  • 1: Rule-based/deterministic logic is sufficient.
  • 5: Requires synthesis of unstructured data and complex multi-step orchestration.

2. The Slog Factor (Operational Pain)

  • The Question: How much human "toil" are we actually removing?
  • Scoring:
  • 1: Minor annoyance or infrequent task.
  • 5: High-frequency (daily/hourly) task with massive hours lost per instance.

3. Business Outcome (CFO Alignment)

  • The Question: Does this tie directly to a metric the CEO/CFO already cares about?
  • Scoring:
  • 1: Marginal improvement to a non-critical process.
  • 5: Critical impact on Revenue Growth, Margin Improvement, or Churn Reduction.

4. Readiness (Data & Process Maturity)

  • The Question: Is the foundation stable, or are we building on "mud"?
  • Scoring:
  • 1: Messy/siloed data and inconsistent tribal knowledge.
  • 5: Clean, labeled data (85%+ accuracy) and documented SOPs.

The Decision Matrix (The "Prague 2026" Thresholds)

Total the points (Max 20).

Total ScoreStrategic DecisionLeadership Action
16-20Invest AggressivelyScale now. This is a "Strategic Win" with high ROI potential.
10-15Narrow Experiment2-week timebox. Prove the "Return on Learning" (ROL) before more funding.
Under 10Kill itDon't let it become a confidence-draining "Science Project."

Use it, adapt it, hand it on. A mention is nice, not required.

Pass it on

Know someone who'd use these?

A message is written for you. Edit it, or send it as it is.

Sunday Field Notes

The talk was one hour. The notes are every Sunday.

One short note most weeks on engineering leadership, teams, and building a career in public β€” the same thinking you just heard, without pretending I've figured it out. No pitch, no cadence guilt, leave in one click.