How to create OKRs

OKRs become far easier to manage when they live in a single, structured Notion database—complete with measurable progress tracking, multiple filtered views, and linked team pages. The core setup treats an Objective as the destination and each Key Result as a quantifiable outcome, then turns every Key Result into its own database entry so teams can attach context like tasks, notes, and ownership while still seeing progress at a glance.

The build starts by creating a table database in Notion and using the first column to store Key Results. Each Key Result entry can be expanded into its own page, where teams can add supporting material such as action items or meeting notes. To make the system usable across an organization, the database then gains properties that function like the columns of an OKR spreadsheet: three single-select fields for Objective, Team, and Quarter; three number fields for Initial Value, Current Value, and Target Value; and a formula property called Progress that calculates completion as a percentage. The transcript walks through concrete examples—employee engagement moving from 0.66 to 0.73 with a target of 0.8, a Glassdoor rating shifting from 3.7 to 3.9 toward 4.5, QA hiring growing from 0 to 2 with a target of 5, and a performance metric where Current Value already matches Target Value once the goal is achieved.

After progress math is in place, additional properties round out operational tracking: a Person property for the lead, a Date property for the time frame (with start and end dates), and a Last Edited Time property to capture when each Key Result was last updated. With these fields, teams can not only measure outcomes but also see who owns them and how recently they were worked.

The second major step adds database views—different ways of looking at the same underlying OKR data without duplicating it. New views can display only one team’s OKRs (e.g., Engineering or People) using filters on the Team property, or show what a specific manager owns using a dynamic filter like Lead contains me. Additional filtered views can isolate a particular quarter, and a Timeline view can lay Key Results out chronologically for a clearer picture of work over time. In total, the system can produce multiple perspectives—seven views in the example—while keeping the source data consistent.

Finally, the transcript shows how to link the OKR database onto each team’s home page using a linked database. That linked version can be filtered and simplified (including hiding properties) so every team sees only what matters—such as Engineering Key Results for a specific quarter—while edits to the underlying OKR entries remain transparent across teams. The result is a single, measurable OKR system that stays visible, focused, and up to date everywhere teams collaborate.