What does it mean that single source of truth per data type?

Each type of information should have one canonical location — avoid duplication.

Example: You keep your task list in three places: a project management app, a notes document, and a paper notebook on your desk. On Monday, you complete a task and check it off in the project management app. On Tuesday, your colleague asks about it and you glance at the paper notebook — where it still appears undone. You tell her it is still in progress. On Wednesday, during a planning meeting, someone pulls up the notes document, which shows the task as active with an outdated deadline. Three copies of the same information, three different states, two miscommunications, and one planning decision made with stale data. Now compare: you designate the project management app as the single source of truth for all tasks. The paper notebook becomes a scratch pad for capturing during the day — items get transferred to the app during your evening sweep, then crossed off the paper. The notes document references the app for task status rather than maintaining its own copy. One canonical location. One current state. Zero drift. The fifteen seconds it takes to check the authoritative source replace the fifteen minutes spent reconciling conflicting versions.

Try this: Conduct a Single Source of Truth Audit for your personal information ecosystem. (1) List every type of information you manage regularly. Common types include: tasks and to-dos, calendar events and appointments, contact information, project notes, reference material, financial records, passwords and credentials, goals and objectives, meeting notes, ideas and captures. Add any types specific to your life. (2) For each type, list every location where that information currently lives — every app, notebook, spreadsheet, folder, or system that contains instances of that data type. Be honest. Check your phone, your laptop, your desk, your email. (3) For each type that exists in more than one location, designate one canonical source. Write it down explicitly: "The single source of truth for [data type] is [location]." (4) For each non-canonical location, decide its role: Is it a capture point (temporary holding before transfer to canonical)? Is it a read-only reference (pulls from canonical, never written to independently)? Or is it redundant (should be eliminated)? (5) For any type where you cannot designate a single source because two locations contain non-overlapping information, plan the merge: what needs to move, where, and by when? (6) Write your SSOT registry — a simple document listing each data type and its canonical home. Post it where you will see it when you are tempted to create a duplicate. Time: 45-60 minutes for the audit, plus migration time for any merges identified.