Flow Metrics Analyzer

Get free access

Tell us where to send updates and the tool opens straight away. It takes about 20 seconds.

Cycle-time percentiles (50th, 70th, 85th, 95th) from your own finished items
Throughput per week, average work in progress and a Little's Law check
A scatter plot, distribution chart and WIP chart you can print or save as a PDF
A forecast: how many items in N weeks, or how many weeks for N items

Your CSV is read in your browser and never uploaded. We only receive the details you enter here.

1. Add your data

You need one row per work item with a start date and a finish date. Items with no finish date are treated as still in progress.

Drop a CSV file here or choose one

The demo is a generated example, not real team data. The tool also accepts exports with a comma, semicolon or tab separator.

How do I export from Jira or Azure DevOps?

Jira: open the issue search, add the columns Key, Created (or the date the item moved to In Progress) and Resolved, then Export > CSV. If you have a status-change date column, use that as the start date, because the created date includes time spent waiting in the backlog.
Azure DevOps: run a query with ID, Activated Date (or Created Date) and Closed Date, then export to CSV. Use Activated Date as the start when it is available.

3. Your flow metrics

Cycle time scatter plot

Each dot is a finished item. The lines show the 50th, 85th and 95th percentiles.

Throughput per week

Items finished each full week. The line is the average.

Cycle time distribution

How many items finished in each range of days.

Work in progress over time

Items started but not yet finished, on each day.

Percentiles

Forecast from your own history

This uses your weekly throughput history and repeats it at random 10,000 times. It is a probability, not a promise.

How many items do you need to finish?

How many weeks do you have?

How these numbers are calculated

Cycle time is finish date minus start date, counting the start day if you left that box ticked. It is in calendar days unless you chose working days.
Percentiles use linear interpolation between ranked values, the same method as Excel's PERCENTILE.INC. The 85th percentile means 85% of items finished in that many days or fewer.
Throughput counts finished items in full Monday to Sunday weeks. A partial last week is left out.
Work in progress counts, for each day, items that had started and not yet finished. Average WIP is the mean of the daily values over the period.
Little's Law says average WIP equals average throughput multiplied by average cycle time, when the system is stable. The check below the cards shows how close your data comes.
The forecast samples your weekly throughput with replacement. It assumes the next weeks look like the past weeks. If your team, process or work mix changed, use only the recent period.

Want to act on what you see?

Flow metrics only help if the team uses them to change how work moves. Our Kanban courses cover how to read these charts, set WIP limits and run a flow review: Team Kanban Practitioner and Kanban Management Professional. AgileFever sells these courses. Additional discounts are available. Contact us to know more.

Written by the AgileFever editorial team. Method notes are in the box above, and the calculations run entirely in your browser. SAFe and Kanban are registered frameworks of their own organisations.