Cycle Time Distribution by Board Column
Each board column has its own panel — for swimlane boards, each column × swim lane combination gets its own panel.
Each blue dot is one closed work item. The Y-axis shows how many days it spent in that column. The X-axis shows the date it exited.
A darker dot means two or more items landed at nearly the same position — they overlap and blend together.
Tip: drag this panel to move it out of the way.
Hollow orange dots are items currently in progress — they haven't closed yet.
Their Y position shows how many days they've already spent in that column.
An orange dot at the right edge is still in this column. An orange dot not at the right edge has since moved to another column. Items high on the right edge are aging in this column and may need attention.
A darker orange dot means two or more in-progress items are at nearly the same position.
Hover any dot to see a tooltip with title, state, cycle time, and more. The same work item highlights simultaneously across all columns and swim lanes — trace its full journey across the board.
Click a dot to open the work item directly in Azure DevOps (opens in a new tab).
The grey shaded band spans from p50 (median) to p85 — the typical range for this column.
Dots inside the band are normal. Dots above the band are outliers — items that lingered longer than 85% of their peers.
Below each column, a WIP sparkline shows how many items were in that column each day during the selected period.
Hover any point on the sparkline to see the exact WIP count for that day.
Spikes signal congestion. High WIP often correlates with longer cycle times — Little's Law in action.
Below each column, a stats strip shows p50 (median — half of items complete within this time) and p85 (85% complete within this time; your worst typical case).
The p85 color signals predictability: blue = predictable, yellow = variable, red = highly unpredictable (p85 ≥ 2× p50).
Hover anywhere in a stats cell to see a fuller description. The item count and outlier count are shown below the percentiles.
Board Lead Time — from item creation to done. Board Cycle Time — from when work actively started to done.
The LT-CT Gap is the difference: time items spent waiting before your team began working on them. A large gap means your biggest win may be reducing queue time, not speeding up the work itself.
Example above: 12d lead time, 5d cycle time → 7 days (58%) was spent waiting, not being worked on.
Right-click any column header in the visual to merge, hide, or reorder that column. You can also copy outliers to your clipboard — by column, by swim lane, or for the entire board.
The Edit Visual tab also shows automatic merge suggestions when columns look like splits of the same stage (e.g. Active — doing / Active — done).
Changes update the visual instantly and are reflected in any AI analysis you run.
Use Save in the Edit Visual tab to persist your edits — they'll be reapplied automatically the next time you load data for the same team.
The Analyze tab sends your flow statistics to an AI model and streams back a tailored analysis.
Answer a few questions about your team and process — work type, SLA targets, team size — to get more specific findings and recommendations.
Only aggregated statistics and your team name are sent — no work item titles are included.
Some columns appear empty — no work items passed through that column during the selected date range.
This is normal for rarely-used columns. Try extending your date range if you expect to see items there.
Edits Explained
Teams naturally change their board configurations over time, leaving a mix of historical and current data in the analytics.
Edits let you adjust the visualization so the historical data reflects your current board structure.
Right-click any column header for the following options:
- Merge into… moves work items from a renamed or deleted column or lane into a current one.
- Hide removes a column or lane from the view - useful for excluding outlier columns like a long-running backlog, or focusing on a specific part of your workflow.
- Move Left/Right reorders columns to match your current board layout.
Suggested Edits detects common patterns like columns split into Doing and Done, and offers a one-click correction.
Applied Edits lists edits currently in effect.
Saved edits are automatically reapplied each time you load the same board. Use Save to keep your current edits or Reset to start fresh.
How to share sanitized data in a bug report
- Set the user selections on the visual to demonstrate the bug.
- Press the generate button then make any other changes necessary to demonstrate the problem.
- Sanitized data is shown in the text areas below, scroll down to review it.
- If you feel the sanitized data exposes something sensitive, uncheck the box in the bug report so your data is not sent and let me know what else needs to be sanitized.
- Describe the issue you see, expected results, repro steps, etc. in the Feedback / Bug Report panel.
- Include your email if you want to be notified when the bug is fixed, email is also helpful if clarification is needed.
- Press Submit to send your bug report, data is only sent if Share sanitized data is checked.
Notes:
- Including sanitized data speeds up problem resolution.
- If you are uncomfortable sharing your data, let me know why so I can improve the data sanitization.
- Your sanitized data is only used to debug, fix, and test your reported problem, and is deleted once resolved.
How your data is sanitized
- ADO Organization is set to “Organization”
- ADO Project is set to “Project”
- Team Name is set to “Team Name”
- Project and team GUIDs are set to “-GUID-”
- Work item titles are set to “Title for id: XXXXXX” where “XXXXXX” is the ID of the work item
- Tag names are set to “TagXX” where “XX” is the order in which the tag was found in the data
- Work item IDs are not sanitized — they are numbers with no meaning outside your ADO organization and project, both of which are sanitized.
- Board column names are not sanitized — they reflect your workflow stages (e.g. “In Progress”, “UAT”). These are rarely sensitive, but if yours are, uncheck Share sanitized data and request additional sanitization.
Large datasets show a preview below for quick review. Copy to Clipboard always copies the full dataset. ⓘ
User Selections
User Edits (0)
Cohorts
Work Items (0)
Revisions (0)
In-Progress Work Items (0)
In-Progress Revisions (0)
Transformed Data (0)
Performance