The OKR cycle

Execution → OKR Scheduler holds one or more named schedule plans, laid out across the four quarters of a chosen year. Each reminder in a plan is a genuinely configurable automated email: Recipient* (the group or user who receives it), Link to page* (which OKR page the email's URL points to), a free-text Message, Frequency* (Yearly, Quarterly, Monthly, Bi-weekly, or Weekly), a Due on the rule for exactly when in that period it's due, and Send reminder — a number of days (or other unit) before that due date the email actually goes out. Reminders sit in two separate lists: Update OKRs — nudging people to keep their own figures current — and OKR retrospective, a second list for scheduling the review meeting itself. Each has its own + Add reminder, but both share the same fields.

Above both lists, a quarter timeline plots every reminder's actual future send dates as small numbered markers, colour-coded per reminder — so a bi-weekly reminder shows up as a dense trail of dots across the year, while a quarterly one shows up once per quarter. It's a preview of the emails that are actually about to go out, not just an abstract cadence setting.

This is what "keeping OKRs current without chasing people" really means in the product: not automation of the numbers themselves, but a scheduled nudge to whoever's Responsible, sent to the right page, on a cadence you set once.

Reading a key result table means reading Progress alongside Status — a status like At risk can appear even when Progress looks reasonable on paper, because Status is always a manual judgement call (9.2) about trajectory and blockers that a percentage alone doesn't capture.