Toolkit

Decision Quality Toolkit

Lightweight decision templates, facilitation recipes, an options matrix, and a decision-type classifier to help teams surface assumptions, compare options, document tradeoffs, and learn from outcomes.

Welcome — Make better decisions that stick

This toolkit gives teams a small set of practical, repeatable tools you can use in meetings and asynchronous work to make decisions faster, clearer, and auditable. Each tool targets a common failure mode: missing assumptions, unclear tradeoffs, forgotten owners, or no learning loop. Use the whole toolkit as a workflow or pick the pieces that match your hunger.

Quick start

  1. Classify the decision using the Decision Type Classifier (section below).
  2. Use the Decision Record template to capture context, options, and the chosen approach.
  3. Run a short options matrix or expected-value worksheet when outcomes and probabilities matter.
  4. Run a pre-mortem for strategic or high-risk choices to surface failure modes early.
  5. Assign an owner and a review date. Log the result and run a brief retrospective after the review date.

1) Decision record template

Use this as the canonical, minimal decision artifact that lives in your knowledge system. Keep each record short — aim for a single page.

Decision Record

  • Title: (short, searchable)
  • Date:
  • Decision maker(s): (names, role or team)
  • Owner (implementation):
  • Decision type: (urgent / reversible / strategic)
  • Context & problem statement: (1–2 sentences describing why this decision matters)
  • Options considered:
    1. Option A — brief description
    2. Option B — brief description
    3. …
  • Key assumptions: (list assumptions that would change the decision if false)
  • Decision rationale & tradeoffs: (why chosen, main pros/cons, critical uncertainties)
  • Measures & signals to watch: (what metrics or events will show success or failure)
  • Implementation plan / next steps: (who does what, by when)
  • Review date / trigger: (calendar date or condition to revisit)
  • Outcome & lessons (filled after review): (what happened, what we'd do differently)

2) Pre-mortem facilitation script & failure-mode prompts

Run a pre-mortem when a decision is important, uncertain, or costly to reverse. Keep it short (15–45 minutes).

Facilitation steps

  1. Frame the decision (2–3 minutes): state the decision and context from the Decision Record.
  2. Imagine failure (5–10 minutes): ask participants to silently imagine the decision failed and write down reasons for the failure.
  3. Share and cluster (10–15 minutes): go around and collect reasons, cluster similar causes, and surface the top 3–5 failure modes.
  4. Mitigate & test (10–15 minutes): for each top failure mode, brainstorm preventive actions, detection signals, and low-cost tests or experiments.
  5. Decide next steps (5 minutes): add actions to the Decision Record with owners and dates.

Failure-mode prompts (use to prime participants)

  • What assumptions would have to be false for this to fail?
  • What could happen outside our control that kills this option?
  • What internal behavior or incentive might subvert the plan?
  • What data would show we're on the wrong path quickly?
  • How might a competitor, regulator, or supplier break this?

3) Options matrix & simple expected-value worksheet

Use an options matrix to compare alternatives on the most important criteria. Use the expected-value worksheet when outcomes can be approximated numerically.

Options matrix (example layout)

Columns: Criteria (cost, speed, impact, risk, dependencies). Rows: Options. Score each option 1–5 against each criterion, note weighting if needed, and compute a weighted total.

Expected-value worksheet (simple)

When you can estimate outcome values (benefit or loss) and probabilities, compute: Expected value = sum(probability * value) for each option. Prefer this when benefits and probabilities are meaningfully quantifiable and different enough to influence the choice.

Example table:

OptionOutcomeProbabilityValueExpected value
Option AHigh adoption0.4$200k$80k
Option BModerate adoption0.6$100k$60k

4) Decision type classifier with recommended meeting designs

Classify each decision to match the process and the degree of documentation needed.

  • Urgent: Needs immediate action. Rule-driven, limited options, little time for analysis. Use a short decision record, assign owner, document the rationale and review date. Meeting design: 10–15 minute huddle or async capture.
  • Reversible: Can be tried and rolled back with low cost. Prefer rapid testing, short experiments, or pilots. Meeting design: small experiment planning session, lightweight decision record, and short review after pilot.
  • Strategic / high-impact: High cost if wrong, many uncertainties, cross-functional implications. Use full Decision Record, options matrix, pre-mortem, and clear review triggers. Meeting design: facilitated workshop, 45–90 minutes, with a documented outcome and owner.

5) Example filled templates (two short scenarios)

Example A — Choose a new helpdesk tool (Reversible)

  • Context: Current tool lacks automation; we need a better SLA workflow.
  • Options: Buy Vendor X, Build minimal internal tool, Use existing plugin.
  • Assumptions: Vendor integrates with our auth system; we can roll back without losing tickets.
  • Decision: Pilot Vendor X for 3 months (owner: IT manager). Signals: integration completed, < 10% SLA violations. Review date: 90 days.

Example B — Open a new regional office (Strategic)

  • Context: Market demand in Region Y; requires lease, staffing, compliance.
  • Options: Full-office launch, small satellite team, remote-first with local partner.
  • Top assumptions: market size, ability to hire required roles, lease cost.
  • Pre-mortem top failure modes: under-hiring, higher-than-expected lease, slower sales.
  • Decision: Start with small satellite team and 6-month pilot; mitigation: hire local recruiter, set break-even metric for month 6. Review date: 6 months.

Using the toolkit in your organization

  • Make the Decision Record the canonical place decisions live (link to implementation tasks, tickets, and measurements).
  • Keep records short and searchable; use tags for decision type, team, and review date.
  • Automate reminders for review dates and require a short outcome entry after review.
  • Run lightweight retrospectives on decisions that exceeded thresholds (cost, time, impact).

Facilitation notes & quick prompts

Short prompts to use aloud or in a chat:

  • "What assumption, if false, would make this a bad idea?"
  • "If this fails, what's the most likely reason?"
  • "Which option gives us the fastest learning with lowest irreversible cost?"
  • "How and when will we know this is working?"

Templates and variants you might add

Examples: a short one-line record for emergency decisions, an executive summary template for board-level choices, an experiment plan template for reversible decisions, or a decision-dashboard that surfaces upcoming review dates and unresolved assumptions.

Use this toolkit as a starting structure — keep what helps and adapt the rest. The goal is predictable, auditable, and improvable decisions, not more paperwork.

Discuss Ask the Librarian

Discussion

Comments and conversation will live here.