When Everything Is Important, Design a Decision Filter

Urgency without a filter turns the founder into the routing system for every request.

Use explicit decision rights and a small filter tied to the current constraint, evidence, opportunity cost and cost of delay.

Write the current constraint; assess five requests; assign or defer; review the filter when the constraint changes.

Focus creates opportunity cost, but unfiltered work creates hidden opportunity cost and context switching.

Can the team explain why a request is being accepted, deferred or declined?

Evidence boundary: SOE editorial decision guide. Evidence informs the framework; it does not establish a universal rule for every business.

Use this on one live decision rather than as a general checklist.

1. Current state: What is happening now, using a small set of observable facts?

2. Decision: What exactly must be decided, and by when?

3. Evidence: Which customer, financial, operational or team evidence would change the decision? Separate evidence from interpretation.

4. Smallest action: What is the smallest reversible step that can produce useful evidence?

5. Owner: Who can make the decision without waiting for the founder?

6. Measure: What will you watch: conversion, contribution, cycle time, quality, cash timing, retention, response or another directly relevant measure?

7. Review point: Put a date on the decision. Do not let an experiment continue simply because nobody defined when it ends.

8. Stop condition: What evidence would make you pause, change or abandon the approach?

• Activity replaces evidence: the team completes tasks but does not learn anything that changes a decision.
• A proxy becomes the goal: a visible number improves while the underlying business problem remains.
• The rule becomes rigid: a useful framework is applied after the context has changed.
• The founder remains the hidden bottleneck: ownership is named but approval still routes back to one person.
• The source is broader than the claim: evidence from one context is presented as if it proves a universal rule.

A strong output from this guide is not a completed checklist. It is a documented decision with an owner, evidence, trade-off, next action and review date. If those five elements are missing, the work is not finished.