Track a decision

A decision worth making is a decision worth tracking. This is the workflow for capturing one properly so you (or whoever owns the work next year) can defend it.

When to log a decision

• The call has real downstream consequences.
• Someone might second-guess it in a quarter or a year.
• It's a pattern you're going to repeat — e.g. "we always reduce spend on channel X when CAC exceeds Y."
• It's tied to a goal you're tracking.

Don't log every choice. Log the ones that matter.

Step 1 — Start from a reasoning answer

The cleanest path: get an answer to the question that informs the decision, then click Save as decision in the answer toolbar.

This pre-fills:

• The question
• The answer (with citations to source versions)
• The assumptions
• The tool trail

You're not re-typing anything that's already on the answer.

Step 2 — Add the actual decision

The answer told you what's true. You decide what to do. Write the decision in one sentence:

• "We will move 30% of Q3 spend from Channel A to Channel C."
• "We will not raise prices this quarter."
• "We will run a churn-save experiment with 200 at-risk accounts."

Step 3 — Add hypotheses

A decision rests on assumptions about the future. Capture them as hypotheses:

• "Channel C ROAS will hold above 2.5x at 30% higher spend."
• "Price elasticity in our segment is < 0.8."

Each hypothesis can be marked validated, invalidated, or unresolved later. This is the loop that feeds your Decision Sign.

Step 4 — List the tradeoffs

Every decision has costs. Listing them up-front:

• Forces you to be honest about what you're giving up.
• Makes a retrospective fair — "we said we'd lose X for Y; did we?"
• Builds your Tradeoff Awareness trait over time.

Step 5 — Define the actions

Break the decision into concrete actions, each with an owner and a deadline. The platform will track which ones complete.

Step 6 — Set the horizon

Every decision has a time horizon — when will we know whether it worked? Set it explicitly. The horizon feeds your Time Horizon trait.

Step 7 — Link a tracker (optional but recommended)

If there's a metric that tells you whether the decision is working, link a tracker. Choose the role:

• Primary — directly measures the decision's outcome.
• Leading indicator — moves before the outcome does.
• Lagging indicator — moves after the outcome does.
• Guardrail — what shouldn't break while you pursue the outcome.

Step 8 — Save

The decision now has a detail page. You can:

• Generate a report to share with stakeholders.
• Re-run the underlying question to see if the answer has changed.
• Mark outcomes as hypotheses resolve.

Why this discipline pays off

• Defensibility. A year later, "why did we do that?" is answered in one screen.
• Pattern recognition. Your Decision Sign becomes meaningful once you have ~20 logged decisions.
• Better next decisions. The platform can surface similar past decisions when you're about to make a new one.
• Hand-off proof. New team members can read the decision log instead of guessing.

What not to do

• Don't log decisions retroactively. It defeats the purpose. Log the decision at the moment you make it.
• Don't skip hypotheses because "it's obvious." It's not, six months from now.
• Don't link a tracker that's already an alert tracker — the goal-linked roles are richer.