Enter observations with sources, dates and limitations.
Compare three options, including doing nothing.
Choose the smallest test that can resolve a meaningful uncertainty.
WORKED EXAMPLE
What a useful result looks like
Fictional example: Northline sells a project-planning app. The next decision is whether to spend a week rebuilding its pricing page. Six recent interviews mention uncertainty about team permissions; the small sample is not representative. The team has no evidence that price caused abandoned trials. Options: rebuild pricing, clarify permissions at signup, or keep the page and interview recent trial users. Choose a permission-copy test first because it is smaller and addresses an observed question. Compare activation in a defined trial cohort, with the same observation window; also count support questions. Choose the evaluation window and threshold before launch. This is a proposed test, with no claimed uplift.
Scope and compatibility
Decision methods and editable instructions. Channel-specific account audits and CRM implementation are separate work.
The files are editable Markdown and CSV. The optional original plugin contains one skill for Codex; its instructions can also be used in a regular AI chat. No accounts are connected automatically.
Want help applying it?
Review your evidence and choose one experiment with Mark. Deliverables and price are agreed before work begins.