Linear Best Practices
Using Linear well is about habits, not features. Every feature in Linear only delivers value when teams apply it consistently and intentionally. This topic brings together the most important practices that high-performing teams use to get the most from Linear every day.
Issue Writing Practices
Write One Issue Per Task
Each issue must describe one and only one piece of work. Combining two tasks in one issue creates confusion — the issue can never be truly "done" until both are done, and the assignee has no clean way to split the work. Break multi-part work into sub-issues instead.
| Wrong (Combined Tasks) | Right (Separate Issues) |
|---|---|
| Fix login bug and update the profile page | Issue 1: Fix login timeout bug on Safari Issue 2: Redesign the profile page layout |
| Improve performance and fix all bugs | Multiple focused issues, each targeting one problem |
Include Acceptance Criteria
Every issue should answer: "How do we know when this is done?" Acceptance criteria remove ambiguity. The developer and reviewer know exactly what the finished state looks like, which prevents issues from bouncing between In Review and In Progress repeatedly.
Backlog Hygiene
A bloated backlog slows everyone down. Teams spend more time reviewing old issues than doing new work. Set aside time every one to two weeks to clean the backlog.
Backlog Cleanup Checklist
| Check | Action if True |
|---|---|
| Issue has been in Backlog for over 90 days | Cancel it or set a realistic future cycle |
| Issue has no assignee and no priority | Review and either assign it or cancel it |
| Issue is a duplicate of another issue | Mark as Duplicate and close it |
| Issue description is empty or too vague | Update with context or cancel if no one remembers why it exists |
| Issue was created by someone who left the team | Reassign ownership or cancel |
Cycle Planning Best Practices
Never Over-Commit a Cycle
Use your team's historical velocity to set the cycle plan. If the average velocity over the last four cycles is 28 points, plan the next cycle for 24–28 points — not 40. Over-committing creates stress, breaks team morale, and inflates carry-over rates.
CAPACITY PLANNING FORMULA Average Velocity = (Sum of last 4 cycles' completed points) / 4 Example: Sprint 20: 30 pts Sprint 21: 26 pts Sprint 22: 28 pts Sprint 23: 24 pts Average: (30+26+28+24) / 4 = 27 points Plan next cycle: 25–27 points maximum Add buffer for interruptions (support requests, meetings): 25 points × 80% buffer = 20 committed points + 5 stretch points
Leave Room for Unplanned Work
Every team gets unexpected tasks during a cycle — urgent bugs, customer escalations, teammate absences. Reserve 15–20% of each cycle for unplanned work. Fill it only when something genuinely urgent arrives, not with extra backlog items.
Workflow State Discipline
Workflow states only reflect reality when everyone moves issues consistently. The most common failure is issues that get stuck in "In Progress" long after work has actually stopped or been completed.
State Ownership Rules
| State | Who Updates It | When to Update |
|---|---|---|
| In Progress → In Review | The assignee | The moment a pull request is opened |
| In Review → Done | The reviewer or team lead | After approving and merging the PR |
| Any state → Blocked | The assignee | The moment a blocker is discovered |
| Done → Cancelled | Team lead or issue creator | Only when the work was never needed |
Priority Management
Keep Urgent Issues Rare
When too many issues carry Urgent priority, the label loses meaning and the team can't determine what truly needs immediate attention. Limit Urgent to issues that actually break production or block users right now. If more than 10% of open issues are Urgent, the priority scale needs recalibration.
HEALTHY PRIORITY DISTRIBUTION Urgent: ▓▓░░░░░░░░ ~5% of issues High: ▓▓▓▓░░░░░░ ~20% of issues Medium: ▓▓▓▓▓▓░░░░ ~45% of issues Low: ▓▓▓▓░░░░░░ ~20% of issues No Pri: ▓▓░░░░░░░░ ~10% of issues
Using Labels Consistently
Define your label set once and write a short definition for each one in a team document. When anyone on the team creates an issue, they should agree on which label applies based on a shared understanding. Inconsistent labels make filters unreliable.
Communication in Issues vs Slack
| Topic | Where to Discuss |
|---|---|
| Technical decision about how to solve a specific issue | In the Linear issue comment thread |
| Quick clarification needing an immediate answer | Slack, then paste the answer back into the issue |
| Design or product debate affecting multiple issues | Slack, but summarize the outcome in each affected issue |
| Status update on a specific issue | In the Linear issue comment thread |
The rule: if a discussion results in a decision, that decision lives in Linear. Slack messages get lost. Linear comments are permanent and searchable.
Daily Linear Workflow
A simple daily rhythm helps individuals stay on top of their work without spending excessive time in Linear.
Suggested Daily Routine
- Morning (5–10 minutes): Open Inbox and process all notifications. Reply to comments, action any mentions.
- Morning (5 minutes): Open My Issues and confirm which issues are In Progress. Move anything that started yesterday from Todo to In Progress.
- During work: Update issue status the moment your work state changes — don't batch status updates at end of day.
- End of day (5 minutes): Add a comment to any In Progress issue noting where you left off. This helps anyone who picks it up next.
Team-Level Practices
| Practice | Frequency | Purpose |
|---|---|---|
| Backlog review | Weekly | Clean stale issues, re-prioritize, assign unowned tasks |
| Cycle planning | Per cycle start | Commit to the right amount of work using velocity data |
| Cycle retrospective | Per cycle end | Review what was completed, carried over, and why |
| Label and workflow audit | Quarterly | Remove unused labels, update workflow states to match real process |
| Integration health check | Monthly | Confirm GitHub, Slack, and other integrations are firing correctly |
