CONFIDENCE WITHOUT EVIDENCEIS JUST LOUD MANAGEMENT.

WHAT COREDECISIONS IS

Records, pillars for trade-offs, metrics on each bet. Manual work: free. Metered automation (e.g. daily tracking) per CreditsCatalog — typically 1 cr per metric per day when that path runs.

HOW IT WORKS

1. 01

   Record the decision

   Calls worth defending — roadmap, monetisation, scope, comms. An artefact the org can return to, not a lost Slack thread.

   
2. 02

   Anchor it to evidence

   Link clusters, insights, snippets, profile — one chain, not three decks.

   
3. 03

   Track daily, for real

   Automated daily tracking: metered per metric per day. Manual paths: no credits. Sustained measurement, not a one-off chart.

   
4. 04

   Close the loop with feedback and reports

   When signal moves, you know which decision it belonged to — Decision & impact snapshot in CoreReports.

   
5. 05

   Worked example — four metrics, one bet

   You ship a live-ops change and attach four automated daily metrics (e.g. recommendation rate, vitriol share, volume on an “economy” cluster, refund-mention signal). While that decision sits in Implemented / Measuring, metering runs at 1 credit per metric per day — about 4 credits/day (~120 over 30 days) for that decision, before any manual edits or one-off report runs. When the window ends, the record still shows what you believed, what you measured, and how the curves moved — so the next room argues from the same evidence chain.

   

CREDITS AT A GLANCE

Reading and manual edits: free. Rows below = metered automation.

CreditsCatalog is source of truth. Stacked metrics example — see Pricing, “What uses credits.”