RunsheetDecision Making

DACI

ComplexityLow

Time30-60 min

Participants2-10

FormatWorkshop + async

MaturityEstablished

Preparation

What needs to be ready before start

Materials

Decision doc template (Notion, Confluence) with sections Driver, Approver, Contributors, Informed, decision question, options, recommendation, decision, follow-up actions; shared area for pre-comments; timer.

People / roles

Driver (drives the decision and collects input); Approver (single person with final decision authority); Contributors (provide input, max 3-5); Informed (kept updated after decision); optional facilitator for complex decisions.

Pre-read

Decision question as one sentence; relevant data, constraints, assumptions; known options or solution spaces; deadline and consequence of no decision; comparable prior decisions (ADRs).

Time needed

30-60 min workshop, plus async lead-in

Setup

Create decision doc, assign roles with names rather than function labels (not "engineering" but specific people). Open doc at least 48h before meeting for pre-comments.

Core question

The one question this method answers

What decision does the Approver make based on Driver preparation and Contributor input, and who gets informed when and by whom?

Flow

Marker: Phase

Step	Duration	Action	Hint
1Phase 1: Assign roles	5-10 min prep	Set roles by name. Ensure exactly one Approver, one Driver, 2-5 Contributors, and an Informed list.	Multiple approvers means no decision. If undecidable, escalate; decision is at the wrong level.
2Phase 2: Driver prepares	1-5 days	Prepare decision doc with question, 2-4 options, pros/cons, recommendation. Contributors read and comment async with at least 48h prep lead time.	Driver writes clear recommendation, not just options. "I'm neutral" means the preparatory job was not done.
3Phase 3: Workshop or decision meeting	30-60 min	Driver presents recommendation. Contributors bring pre-comments into 15-20 minute discussion. Approver asks clarifying questions.	The session is not for defining the problem from scratch. If someone has not read the doc, they cannot contribute comments. Discipline drives efficiency.
4Phase 4: Approver decides	5-15 min	Approver states decision with reasoning. Add dissent note if needed (contributors who disagree with short argument).	Disagree and commit: contributors may dissent, but once decided they commit to implementation. Dissent note protects diversity.
5Phase 5: Documentation and communication	1-2 business days	Driver finalizes doc and sends communication to Informed (often channel or mail). Record follow-up actions with owner and date.	Without communication to Informed, decision spreads by rumor. Set channel and format in advance.

Artifact

What comes out at the end

Form

Decision document with header (question, date, status), role list with names, options with pros and cons, recommendation, decision with rationale, optional dissent note, follow-up actions with owner and date, Informed distribution list.

Versioning / ownership

Decision doc is immutable after decision; later revisions become a new entry linking to the old one and explaining status change.

Tool alternatives

Notion or Confluence page with DACI template
Markdown ADR variant in repo
Atlassian decision macro
Linear project with Decision workflow

daci-working-template.md

Compact working template for DACI with roles, options, criteria, and decision log.

DACI Working Template

Goal

Clarify who drives the decision, who approves it, who contributes, and who stays informed.

Context

What decision needs to be made, and why now?

Input

Decision question:
Stakeholders:
Constraints:
Facts and assumptions:

Working area

Driver:
Approver:
Contributors:
Informed:
Options:
Criteria:

Output artifacts

DACI matrix:
Decision log:
Follow-up actions:

Open questions

Decision / next step

Owner, date, and success signal.

Example output

Concrete filled scenario, fictional example

daci-beispiel.md

Concrete filled scenario, fictional example

DACI — switch from Heroku to AWS for backend hosting (Decision date: 17.05.2026)

Roles

Driver: @anna (lead platform)
Approver: @marcus (CTO)
Contributors: @ben (SRE), @lisa (backend lead), @julia (finance director)
Informed: engineering org (15 people), customer success lead, security officer

Question: Do we switch to AWS (EKS + RDS) by Q4/2026?

Options

AWS EKS + RDS, 4-month migration. Estimated one-time cost 95k EUR, monthly -3.2k EUR from month 5.
Stay on Heroku and upgrade to Performance Dynos. Monthly +2.1k EUR, no migration.
Hybrid: critical services on AWS, rest on Heroku. Complex and harder to maintain.

Pros/cons summary: Option 1 is cheaper and more controllable long term but riskier to migrate. Option 2 keeps status quo, no learning progression. Option 3 is hard to maintain.

Driver recommendation: Option 1.

Pre-comments: @ben +1 (experience from Team Discovery). @lisa expresses timeline concerns. @julia requests sensitivity table for costs.

Approver decision: Option 1 with condition: migrate in 5 months instead of 4, plus risk buffer. Extend cost model before migration.

Dissent: none.

Follow-up actions:

Migration plan v1 (Owner: @anna, by 30.05.).
Cost model (Owner: @ben, by 24.05.).
Communication to Informed (Owner: @anna, by 19.05.).

What to do

Informed list is mandatory in template. Treat communication as required follow-up with deadline.

Stop criteria

Done signals checkable in under a minute

Approver is not decision-authorized or not available, so decision must escalate.

Question is too small (operational), DACI overhead is not justified.

Question is strategic and needs broader consensus building, so Consensus Workshop or RAPID is more appropriate.

Driver and Approver are the same person and separation is not possible.

Pre-read discipline cannot be maintained, meetings run without preparation.

Question is mainly political; factual DACI will not solve the issue.

Finished the runsheet?

Go to the profile for purpose, similar methods, and sources or continue to the next method in the catalog.

Go to profile More methods