Quarterly Planning (OKR to Backlog)

Why the quarter opens with capacity, not commitment

The template opens with six subtasks containing no product work at all. Quarter Kickoff & Alignment runs on days 0 to 3 — finalize OKRs with leadership, present them to engineering and design, identify cross-team dependencies, complete a resource allocation review — and the first key result does not open until day 5. That gap is the argument of the whole template.

A quarter you commit to before you know what you have is a forecast wearing a plan’s clothes. The day-3 resource allocation review is where the quarter’s real working days get counted. Ninety calendar days is roughly sixty working days, and then you subtract public holidays, booked leave, the on-call rotation, and the share of every week that disappears into support. Skip that subtraction and you find out in week two, when the first KR slips and nobody can say whether the estimate was wrong or the capacity was never there.

The offsets here count calendar days. To count working days instead, the template editor accepts business-day offsets — [Start + 5bd] — and a date variable can skip weekends and a region’s public holidays.

Why you size before you sequence

Every key result opens with measurement, not build. KR1 starts with “Analyze current activation funnel data” on day 5 and “Identify top 3 drop-off points” on day 7; the implementation subtask does not appear until day 28. KR2 measures time-to-first-value before anyone designs its replacement.

A key result stated as “40% to 55%” is a target, not a plan. Until the funnel data says where people actually drop off, anything scheduled against it is a guess with a date attached — and the quarter’s engineering capacity goes to the experiment someone thought of in the kickoff meeting rather than the one the data justifies. The three-week gap between analysis and implementation in KR1 is not slack. It is the design loop.

Who owns which gate

• Leadership owns the objective. “Finalize OKRs with leadership” is day 0 because a quarter whose goals are still being negotiated in week three has no week three.
• @product_lead owns the quarter as an object — kickoff, the day-45 review, the at-risk call, the score. This role never owns a KR's delivery, which is what keeps the mid-quarter review honest.
• @eng_manager owns cross-team dependencies and resource allocation. The one gate the PM does not hold, deliberately: the person who knows who is on-call, who is on leave, and which team owes you an API is the engineering manager, not the roadmap.
• @product_manager owns each KR end to end — dashboard, monthly checkpoints, A/B test, the beta gate at 10% of new signups on day 56, full rollout on day 70.
• @analyst owns the baseline on day 5 and the final snapshot on day 85. Same role at both ends, so the number that scores the quarter is measured the way the opening number was.
• @designer, @dev_team, @qa_team own delivery inside a KR, not the KR itself. Nobody here is asked whether the quarter is on track.

What breaks when you skip a phase

• No resource allocation review. The plan is wrong by week two and stays wrong, because missing capacity is invisible — it reads as the team underperforming rather than the quarter being oversubscribed from day one.
• No dependency scan. Day 2 is early enough to change someone else’s quarter. Week six is not: their plan is committed and yours has a hole in it shaped like their API.
• No baseline. “Score each KR (0.0 to 1.0)” on day 86 needs the day-5 number. Without it, scoring becomes an argument about whether the quarter felt productive.
• No mid-quarter review. Day 45 exists so re-prioritisation happens while six weeks remain. Find a dead KR in week eleven and the only move left is writing it up.
• No retrospective. The load-bearing line is the last one, “Draft next quarter OKR candidates” on day 90 — the first thing dropped when a quarter ends badly, which is when it is worth most.

Five people versus fifty

A five-person team plans a quarter in an afternoon, and much of this ceremony is overhead at that size. Drop the tracking dashboard — three key results, everyone can recite them. Drop the weekly check-in; you talk daily. Drop the cross-team dependency scan, because the cross-team dependency sits two desks away.

Three things still earn their place at five: the baseline, the mid-quarter review, and the score. Small teams are the likeliest to finish a quarter unable to say whether the work moved anything, because nobody wrote down what it looked like before.

At fifty, the parts a small team skips are the entire job. Dependencies become queues with lead times. The mid-quarter review is where four teams discover they each assumed the other three would ship first. The dashboard exists because no one person sees all three KRs any more. The realistic shape is one of these per team plus a parent epic carrying the objective and the checkpoint dates.