How to Set Up Cycles in Linear (Duration, Cooldown, and Rollover)
Cycles are Linear's version of sprints, but they behave differently from the sprints most teams know from Jira. A cycle runs on a fixed schedule whether or not anyone 'starts' it, unfinished issues can roll forward automatically, and there is no ceremony of closing a sprint by hand. That design removes a lot of admin work, but it also means the decisions you make in team settings on day one shape every planning conversation that follows.
This guide walks through enabling cycles for a team, picking a duration and cooldown that fit how your team actually ships, configuring auto-add and rollover behavior, and using the cycle view to spot scope creep mid-cycle. It is written for PMs and engineering leads who own the planning rhythm and want cycles to reflect reality rather than wishful thinking.
Planning a sprint in Linear? Turn your spec into dependency-ordered tickets.
Paste the feature or spec. Vantage pulls out the requirements and breaks them into tickets, ordered by what blocks what, ready to groom with your team.
How this works
- We save what you typed. No AI runs on this page.
- You create a free account.
- Vantage builds a PRD and dependency-ordered tickets from it inside your own workspace.
Why ask for an account? We’re a small team, and a full generation runs a multi-step AI pipeline that costs us real money. So we tie it to a free account instead of running it anonymously. You also get to keep the result and edit it.
Your text is only used to create your project, and it stays private to your account. If you don’t sign up, we delete it after 7 days. Privacy policy
Step-by-step guide
Enable cycles in team settings
Cycles are configured per team, not per workspace. Open Settings, pick the team under Teams, and go to Cycles. Turn cycles on. Because each team has its own schedule, a platform team can run two-week cycles while a growth team runs one-week cycles. Decide this deliberately: if teams frequently hand work to each other, aligned start days make cross-team planning much easier.
- Open Settings and select the team you want to configure
- Open the Cycles section and enable cycles
- Check whether other teams already run cycles and note their start day
- Agree with adjacent teams on whether to align schedules
Choose a cycle duration and start day
Linear lets you pick the cycle length in weeks and the day the cycle starts. Two weeks is the common default, but one-week cycles work well for teams shipping small increments with frequent releases, and longer cycles suit teams with heavier integration or QA work. Pick a start day that does not collide with your release day; starting on Monday and releasing on Thursday or Friday gives a clean planning-then-ship rhythm.
- Pick a duration that matches how long a typical feature slice takes to ship
- Choose a start day that avoids your release and on-call handoff days
- Set how many upcoming cycles Linear should create in advance so you can plan ahead
Decide whether to use a cooldown
A cooldown is a gap between cycles where no cycle is active. Teams use it for bug fixing, tech debt, exploration, or simply to plan the next cycle without the current one bleeding into it. If your team constantly ends cycles with half-finished work and starts the next one already behind, a short cooldown is often a better fix than shorter cycles. If you skip the cooldown, schedule planning inside the last day or two of the cycle instead.
Configure auto-add and rollover behavior
In the same settings, Linear offers options to automatically add issues to the active cycle when they move into a started or completed state, and to move unfinished issues into the next cycle when a cycle ends. Auto-add keeps the cycle an accurate record of what the team actually worked on, including unplanned work. Rollover saves manual cleanup, but it can hide chronic overcommitment, because nothing forces a conversation about why an issue slipped twice.
- Enable auto-add for started issues so unplanned work shows up in the cycle
- Enable rollover of incomplete issues if you want Linear to carry work forward
- Agree on a rule for issues that have rolled over more than once, such as re-scoping or moving back to the backlog
Plan the next cycle from the backlog
Open the upcoming cycle from the team's Cycles page and pull issues in from the backlog. Use estimates if your team has them enabled; Linear shows the total scope of the cycle so you can compare it with what the team completed in recent cycles. Keep the cycle anchored to one or two outcomes, and resist filling every slot. A cycle loaded to 100% of past throughput leaves no room for bugs or review cycles.
- Filter the backlog by priority and project to find candidate issues
- Drag or assign issues into the upcoming cycle
- Compare total estimated scope to what the team completed in the last few cycles
- Leave headroom for bugs, review feedback, and interrupts
Read the cycle graph during the cycle
The cycle view shows a graph of scope against started and completed work over time. A rising scope line mid-cycle is the clearest signal of scope creep. A flat completed line for the first half of the cycle usually means work is too large or blocked in review. Check the graph at standup a couple of times a week rather than only at the end, when it is too late to adjust.
Review the cycle when it ends
When a cycle closes, look at what was completed, what rolled over, and what was added after the start. These three numbers tell you more than velocity alone. If added scope is consistently high, your intake process needs work. If rollover is consistently high, the team is overcommitting or issues are too large. Use the review to adjust the next cycle's commitment, not to assign blame.
Common mistakes
Treating rollover as free
Automatic rollover is convenient, but an issue that rolls over three cycles in a row is a planning problem wearing a disguise. Track repeat rollovers and force a decision: split the issue, deprioritize it, or unblock it.
Changing the cycle length every few weeks
Teams that keep adjusting duration lose the ability to compare cycles. Pick a length, run it for at least a quarter, and only then change it based on what the data shows.
Planning to 100% of past throughput
Past completed scope already includes some interrupts, but not the bad weeks. Plan to roughly 70-80% of your recent average so the cycle survives an incident or a sick day.
Tips
Align cycle start days across teams that share dependencies so cross-team asks land at the start of a cycle, not the middle.
Use a label like 'unplanned' on issues added after the cycle starts so you can quantify interrupt load over time.
Write a one-line cycle goal in the cycle description so standups have something to measure progress against.
Review the upcoming cycle two days before it starts so blockers surface while there is still time to resolve them.
How Vantage helps
Cycles work best when the issues going into them are already scoped and sequenced. Vantage generates tickets from your PRD with dependencies mapped and work grouped into waves, so the first wave is a natural candidate for the next cycle. Push them to Linear and bi-directional sync keeps status changes from the cycle flowing back into the spec.