Decision Record & Rationale Template (ADR-style)

Purpose

Use this lightweight decision record to make important decisions traceable, transparent, and learnable. The template helps teams surface assumptions, compare options, document trade-offs, assign owners, and plan follow-up checks so decisions improve over time and knowledge accumulates.

When to use this template

• Decisions that affect multiple teams, customer outcomes, architecture, process, policy, or budget.
• Choices likely to be revisited, cause downstream work, or carry measurable risk.
• When you want a short, shareable record that supports retrospective learning.

How to use

Fill the template during or shortly after the decision meeting. Keep each field concise—this is a living record, not a long memo. Link to evidence and follow through on the review date and indicators you define.

Template fields (with guidance)

1. Title

   Short descriptive title. Example: "Select default image storage for Product Catalog"

2. Decision date

   Date the decision was made (or the record was approved).

3. Context & background

   2–5 sentences explaining why the decision matters now and any relevant constraints (customers affected, timelines, compliance, cost limits).

4. Stakeholders

   People, roles, or teams impacted or consulted. Include contact person for questions.

5. Choices considered (with pros/cons)

   List each viable option and 2–4 short pros/cons that materially influenced the choice. Keep this evidence-linked where possible.

6. Chosen option

   What you decided and the essential reason(s) for choosing it.

7. Owner & review date

   Who owns implementation and who will review the decision on a specific date (or cadence). The review date should check assumptions and leading indicators.

8. Assumptions & unknowns

   Key assumptions that must hold for the decision to succeed and important unknowns to reduce via experiments or monitoring.

9. Leading indicators to watch

   Concrete, measurable signals that tell you early whether the decision is having the intended effect (e.g., latency < 200ms, adoption rate > 20% in 3 months, defects <= 0.5%).

10. Related experiments or mitigations

    Small tests, rollouts, feature flags, or mitigations planned to reduce risk or validate assumptions. Link to experiment records.

11. Evidence & links

    Links to designs, benchmarks, cost estimates, vendor quotes, user research, meeting notes, or issue trackers used to inform the choice.

12. Outcome & follow-up notes (to be filled after review)

    Record what happened, whether indicators met expectations, what you learned, and any changes to policy or process.

Short facilitation script (5–15 minutes)

1. State the decision question and desired outcome (30–60 sec).
2. Quick context recap: constraints, timeline, and stakeholders (1–2 min).
3. Read options considered and highlight key pros/cons (2–4 min).
4. State the chosen option and the primary reason(s) (1 min).
5. Assign owner and a review date; list 1–3 leading indicators and one experiment or mitigation (1–2 min).
6. Confirm where this record will be stored and who must be notified (30–60 sec).

Worked example (brief)

Title: Default image storage for Product Catalog

Decision date: 2026-05-10

Context: Product images are currently stored on a legacy CDN that charges per request. Increased product views trip monthly costs. Need a cost-predictable option that integrates with our build pipeline.

Stakeholders: Product, Platform, Finance; contact: @alice

Choices considered:

• S3 + CloudFront — Pros: predictable storage, cheaper at scale; Cons: engineering effort to migrate, need caching rules.
• New managed CDN plan — Pros: easy to migrate; Cons: higher monthly fees, limited custom controls.
• Keep legacy CDN but add cache layer — Pros: minimal migration; Cons: may not fix cost model long-term.

Chosen option: Migrate canonical images to S3 + CloudFront to reduce per-request costs and enable lifecycle policies. Chosen because long-term TCO and integrations outweigh migration effort.

Owner & review date: Platform lead (@sam); review 2026-08-01

Assumptions & unknowns: Assumes migration script will handle metadata consistently; unknown: exact cost savings in Q1 post-migration.

Leading indicators: Average per-month image delivery cost drops by 30% within 90 days; cache hit ratio > 85%.

Related experiments: Pilot migrate 10% of catalog and measure costs and latency for 30 days.

Evidence & links: Migration plan (link), cost model (link), pilot playbook (link).

Practical tips

• Keep records short and link to deeper evidence rather than embedding long documents.
• Use review dates as non-negotiable checkpoints—don’t let the record gather dust.
• Prefer 2–3 leading indicators that are measurable and owner-responsible.
• When a decision is reversed, record why: reversal notes are high-value learning artifacts.

Adoption suggestions

Store decision records where teams already look for guidance (team wiki, decision repository, project folder). Encourage teams to link decisions from relevant tickets, PRs, or policy pages so the rationale travels with the work.

Next steps for teams using this template

1. Try the facilitation script in your next decision meeting.
2. Keep 1–2 recent decisions visible in the team onboarding or handbook page.
3. At each review date, add outcome notes to the record and update stakeholders.