Drop your CSV here or click to browse
Works with any CSV containing workflow stage date columns · Data stays on your machine
Column mapping
Change these anytime — charts update instantly
Required
Commitment point — when the item entered your flow, e.g. "To do", "Committed"
When the item was done — e.g. "Closed", "Released"
Where WIP counting begins — e.g. "Doing". If different from Start, items between Start and here show as committed queue, not WIP.
Optional
Ticket reference for tooltips
Display name for tooltips
Ticket URL — click dots on charts to open in a new tab
Days the item was blocked — used in flow efficiency
Boolean flag — Yes/No, True/False, 1/0. If not mapped, blocked indicators are hidden.
Filters
Rows combine (AND) · hold Cmd/Ctrl to select multiple values in one row (OR)—
Items completed
—
Cycle time p50
—
Cycle time p85
—
Throughput avg/wk
—
WIP (today)
—
Flow efficiency
—
Items blocked
Speed
Cycle time is measured from your configured start stage (e.g. To Do) to your end stage (e.g. Done). It is the total time an item spent in your system — including queues. This is what your stakeholders experience as "how long did that take?"
The 85th percentile is your Service Level Expectation — commit to delivering 85% of items within this number of days. It's realistic (not best-case) and builds trust over time.
How fast do we finish?
Cycle time percentiles — how long completed items actually took
No data yet
Forecast
Monte Carlo simulation runs your historical weekly throughput through 10,000 simulated 30-day periods to estimate how many items you are likely to complete. The result is a probability distribution — not a single number.
How to read it: "85% chance of 6 or more" means that in 85% of simulated futures you complete at least 6 items. More reliable than velocity because it captures the natural variability in your actual delivery, not an idealised average.
What can we close this month?
Monte Carlo — likely completions in the next 30 days
No data yet
WIP
Work in progress = items that have entered your start stage but not yet reached your end stage as of today.
Little's Law: Lead Time = WIP ÷ Throughput High WIP directly causes long lead times. Reducing WIP is usually the fastest lever for improving delivery speed.
How much are we juggling?
Items currently in progress
—
Risk
Smelly tickets are WIP items older than your 85th percentile lead time. By definition, 85% of your completed items finished faster than this threshold — so anything still in progress beyond it is a statistical outlier worth reviewing.
The name comes from code smell — not necessarily broken, but worth a closer look. Ask: is it blocked? Forgotten? Too large to finish? The older it gets, the more it is slowing everything else down (Little's Law).
Smelly tickets
WIP items in progress longer than your 85th percentile lead time
No data yet
Focus
Multitasking index = for each item that was started, count how many other items were already in progress at that moment. Average this across all items. A high number means the team is constantly context-switching.
Research suggests that switching between more than 2–3 tasks simultaneously causes significant overhead. Each switch costs focus time and delays everything in flight. High multitasking index is a leading indicator of long cycle times.
How to improve it: Set a WIP limit. Finish existing items before starting new ones. Even one person fully finishing one thing before picking up another makes a measurable difference.
How many balls in the air when we start something?
Average in-progress items at the moment each new item is started
No data yet
Are we finishing what we start?
What this shows — items started (entered your start stage) vs items completed (reached your end stage) each week. When the green "completed" bars are consistently higher than blue "started" bars, your backlog is shrinking. When started consistently outpaces completed, WIP grows and cycle times will lengthen.
What to look for — a healthy team finishes roughly as much as it starts each week (steady state). A team that repeatedly starts more than it finishes is accumulating WIP debt — Little's Law guarantees this will show up as longer cycle times. Spikes in "started" without matching spikes in "completed" often signal unplanned work or poor commitment discipline.
The commitment point matters — if your start column is your backlog (not your commitment point), this chart will show large volumes of items "starting" that haven't actually been committed to. Map your start column to the column where work is actively pulled, not where it's first created.
Items started vs completed per week
Are tickets getting stale?
What this shows — WIP count (green bars, left axis) and average age of in-progress items in days (purple line, right axis) over time. Rising bars mean more work is being started. Rising average age means items are sitting longer before completing.
Average age is the key signal — if it's climbing while WIP count is flat, the same items are ageing in place and nothing is finishing. That's a flow problem. Use the Aging WIP tab to find exactly which items are driving it.
Healthy pattern — WIP count is flat or gently declining, average age is flat or declining. This means the team is completing work at least as fast as new work arrives.
WIP count (bars) and average age in days (line)
Lead Time Control Chart
Each dot is one completed item. Percentile lines show where most of your work lands — aim to keep items below the 85th.
What is cycle time?Cycle time is measured from when active work begins (your active stage) to when the item is done. Lead time (commit→done) covers from when the team commits to doing the item through to completion — this is what the customer experiences as wait time. If no separate active stage is configured, cycle time and lead time are the same.
What do the percentile lines mean?50% of items complete within the green line. 85% complete within the orange line. Use the 85th as your Service Level Expectation (SLE) — a realistic commitment: "We will deliver 85% of items within X days."
Orange dots were blocked at some point during their journey. Compare blocked vs unblocked dots to see how much blocking inflates lead time.
What to look for: Dots clustering tightly below the 85th line = predictable team. Outliers above the 95th = worth a conversation. Trend upward over time = system is slowing — check WIP.
50th percentile
70th percentile
85th percentile
95th percentile
Throughput
Items completed per week. The orange line is a 4-week rolling average — flat and steady is what you want.
What is throughput?The number of items completed in a given week. This is the engine of your system — throughput drives forecasts, and combined with WIP it determines lead time (Little's Law).
Why use a rolling average?Weekly counts are noisy (holidays, sprints, batch releases). The 4-week rolling average smooths this out so you can see the real trend.
What to look for: Steady or rising average = healthy. A sustained drop = investigate (team capacity, blocking, process issues). Spiky pattern with long gaps = work is being batched rather than flowing continuously.
Throughput and forecasting: The Monte Carlo forecast on the Flow Insights tab samples this historical throughput distribution to estimate future completions. More stable throughput = tighter, more reliable forecast range.
Normal range (toggle): Tick "Show normal range" to overlay a band showing your expected weekly throughput based on actual historical variation. Bars outside the band (shown in orange) represent weeks where something real changed — not just random ups and downs. The panel below the chart lists those weeks and whether they were unusually high or low.
— orange bars fell outside your normal variation and are worth investigating
Throughput Distribution
How often the team completed X items in a week. A single narrow peak = consistent, forecastable delivery. A wide spread or two humps = high variability.
What this shows — the shape of your weekly throughput. Each bar answers: "how many weeks completed exactly X items?"
Why it matters — Monte Carlo samples this distribution to build forecasts. A tight peak means reliable forecasts. A wide spread or bimodal shape (two clusters) produces a wide confidence range — the data itself is uncertain, not just the forecast.
What to look for: A single peak near your average = healthy. Two separate clusters = something changed mid-period (team capacity, scope mix). Long right tail = occasional high-output weeks pulling up your average.
Team Health
A scorecard comparing the current period against the previous period of the same length. Green = improving or stable. Amber = watch. Red = take action.
What this shows — key flow metrics for the selected period, compared to the same-length period before it. Each card shows the current value, the direction of change, and a traffic light signal.
Throughput — items completed per week. Higher is better. Declining throughput is the first sign of a flow problem.
Lead time (p50) — median time from commitment to done. Lower is better. Rising lead time usually follows rising WIP — Little's Law in action.
System WIP (avg/wk) — average number of committed-but-not-done items per week across the period. This is a historical average, not today's live count — it will differ from the WIP number in the summary bar, which shows current items only.
Backlog size — items in pre-commitment stages (before your Start column). Only available if you have stages configured before your Start column. Growing backlog means demand is exceeding commitment rate.
FFE time (p50) — median time items spend in the backlog before commitment. Only available with pre-commitment stages. Rising FFE means decisions are taking longer — a prioritisation or capacity issue upstream of your flow.
Flow efficiency — percentage of lead time spent in active work vs queues. Requires stages to be marked P (process) or Q (queue) in Config.
Important: If your Start column is where work first enters the system (not where the team commits to it), backlog and FFE metrics cannot be separated from system metrics. Learn about the commitment point →
Period
All metrics below cover the selected period only. The summary bar at the top uses your full date-filter range — numbers will differ.
Work Split
How your work breaks down by type — across each stage of the pipeline. Compares what's waiting, in flight, and done.
What this shows — the composition of your work at each stage: Backlog (not yet committed), Committed (committed but not actively in progress), In Progress / Active WIP, and Done. Each bar shows how many items of each type sit in that group.
What to look for — if your In Progress bar is dominated by one type (e.g. bugs / failure demand) but your Backlog is mostly features, the team is being pulled into reactive work. If Done is heavily weighted to one type, that's where your effort has actually gone — which may or may not match intent.
Split by — choose any categorical column from your data. Demand Type, Demand Source, ticket type or team are all good choices. Charts with many distinct values can get busy — fewer categories read more clearly.
Committed group — only appears if you have an Active Work Starts column configured. Items committed (Start date) but not yet actively worked appear here. Without that column, all in-progress items appear as a single WIP group.
Backlog group — only appears if you have pre-commitment stages configured before your Start column in the workflow setup.
Breakdown
Key metrics split by a column of your choice — lead time, throughput, blocked rate, and WIP per category.
What this shows — a side-by-side comparison of flow metrics across categories (e.g. ticket type, demand type, priority). Useful for identifying which categories take longer, are blocked more often, or have higher WIP.
Use this unfiltered. The point of this tab is comparison — if you filter to a single category, the table collapses to one row. Apply slice filters on the other tabs to drill in once you've identified which category to investigate.
Exclusion filters are respected (e.g. "hide Not Done") — those apply everywhere to keep the data clean. Slice filters ("show only Priority: High") are ignored here because they defeat the comparison.
LT p50 / p85 — suppressed (shown as —) for categories with fewer than 5 completed items, since percentile stats are unreliable on small samples.
WIP Over Time
Work in progress count each week. A stable or declining WIP leads to more predictable flow — rising WIP means more is starting than finishing.
Little's LawThe most important equation in flow metrics: Lead Time = WIP ÷ Throughput If WIP doubles and throughput stays flat, lead time doubles. You cannot improve lead time without managing WIP.
Rising WIP means more is being started than finished. This is the single most common cause of long lead times — not team capacity or skill, but too many things in flight at once.
What to look for: Flat or declining WIP = sustainable system. Steadily rising WIP = overloaded system heading for slowdown. A sudden WIP spike = something got stuck (check Aging WIP for the culprit).
What to do: Set a WIP limit. Stop starting, start finishing. Even a soft limit ("flag it if WIP goes above X") creates useful conversation.
Lead Time Over Time
Weekly median lead time — rising means delivery is slowing down, falling means it's improving.
What to look for: A flat or falling median line = your system is stable or improving. A rising line = work is taking longer — check WIP Over Time to see if rising WIP is the cause (Little's Law: CT = WIP ÷ Throughput).
p85 line — the lead time 85% of items beat that week. Watch the gap between median and p85: if it widens, outliers are getting worse even if the median looks fine.
Sparse weeks — weeks with very few completions produce noisy CT values. The 4-week rolling median smooths this out; trust the trend line more than any single week.
Weeks with no completions are shown as gaps — this is normal for quiet periods or holidays.
days
Cumulative Flow Diagram
Done anchors the bottom — its slope is your throughput. To Do sits at the top — its growth is your commitment rate.
How to read the bandsEach coloured band is a workflow stage. The height of the band at any point = WIP in that stage. A band that widens over time = work is accumulating there — a bottleneck. A flat or shrinking band = flow is moving through.
Slopes and linesThe slope of any line = its throughput rate (items per day). Two lines with the same slope = balanced flow between those stages. The horizontal distance between any two lines at a point = approximate cycle time between those stages.
Stage snapshot panelHover over the chart to see WIP per stage for that week. The ~Xd is an approximate cycle time for that stage using Little's Law: CT ≈ WIP ÷ throughput rate It reflects flow at that point in time — not the same as historical average stage time.
What to look for: Parallel lines with equal slope = smooth, predictable flow. Diverging lines = WIP growing. Flat bottom (Done) line = throughput has stalled. A big orange or red band = your bottleneck.
Focus modeWhen a large Done count makes the in-progress bands hard to see, click Focus: Off to turn Focus on. It raises the Y-axis floor to the current bottom-line value and trims the date range to match — so only the visible, active bands fill the chart. Use the week dropdown next to the button to control how much history is shown (default 4 weeks). The note below the controls tells you what period has been hidden. Toggle off any time to return to the full view.
Band height = WIP in that stage
Band widens = work piling up, possible bottleneck
Line slope = throughput rate
Horizontal distance between two lines = cycle time
Flat line = no flow through that stage
Zoom:
—
Hover over the chart to see stage breakdown
Click chart to place a marker · click again to add a second
* Cycle time approximated using Little's Law (WIP ÷ recent throughput)
Lead Time Distribution
How often items take a given number of days. A tight cluster to the left = predictable team. Long tail to the right = outliers worth investigating.
What this showsThe shape of your delivery. Each bar = how many items completed at that lead time. The shaded background regions correspond to percentile bands (50th, 70th, 85th, 95th).
Tight cluster on the left means most work completes quickly and consistently — highly predictable. The team has found a reliable way of working.
Long right tail means some items take much longer than most. This makes forecasting harder and erodes stakeholder trust. Usually caused by: items being too large, blockers, context switching, or items sitting in queues.
Two peaks (bimodal) suggests two different types of work with different natural lead times — consider whether they should be tracked separately.
Flow Efficiency Distribution
How much of each item's total time was spent in active (process) stages vs waiting in queues or blocked. High 100% bar usually means items went straight from open to done — filter it out to read the real pattern.
What is flow efficiency? Active time (stages marked P) divided by total time including queues (stages marked Q) and blocked days. 100% means an item moved continuously with no waiting — in practice this often means dates were entered in a single click rather than reflecting real activity.
Why filter out 100% (or near-100%)? Automation and manual date cleanup inflate the top bar and compress everything else. Hide those items to see where real work actually spends time waiting.
What good looks like: Most flow metrics teams aim for 20–40% flow efficiency — the majority of lead time is wait time, not active work. Higher is better, but a realistic average is more useful than one inflated by data quality issues.
Stage types matter: Mark each stage P (process) or Q (queue) in the workflow configuration above. Queue stages and blocked days count as wait time.
%
Aging WIP
Items in progress by current workflow stage and age. Colour shows where each item sits against your completed lead time percentiles.
What this showsEvery dot is a work item currently in progress. X-axis = which stage it is currently in. Y-axis = how many days it has been in the system (age from start date to today).
Colour = risk level compared to your historical lead time percentiles:
🟢 Green — below your 50th percentile. On track, no concern.
🟡 Yellow — between 50th and 70th. Worth keeping an eye on.
🟠 Orange — between 70th and 85th. At risk. Consider a conversation.
🔴 Red — above 85th percentile. 85% of completed items finished faster than this. Needs attention now.
🟢 Green — below your 50th percentile. On track, no concern.
🟡 Yellow — between 50th and 70th. Worth keeping an eye on.
🟠 Orange — between 70th and 85th. At risk. Consider a conversation.
🔴 Red — above 85th percentile. 85% of completed items finished faster than this. Needs attention now.
Click any dot to see the ticket details below the chart. Click "Open ticket →" to open it in a new tab.
What to look for: Red dots in early stages (To Do, Doing) are most concerning — they have a long way to go. Red dots in Done but not yet closed = process issue, not delivery issue.
Fuzzy Front End & Lead Time
How long items wait in the backlog before commitment, and total lead time from creation to completion.
Fuzzy Front End (FFE) is the time from when an item enters the backlog to when the team commits to doing it (your Start column). This is the prioritisation and selection phase — it reflects how quickly your team makes decisions, not how fast it delivers.
Commitment point is the moment the team signals intent to work on an item. Lead time is measured from this point to done. Items can sit in the backlog for months before commitment — including that time in your lead time figure hides delivery capability. FFE (Fuzzy Front End) captures that pre-commitment phase separately.
Total time = FFE time + lead time (commit→done). FFE is the pre-commitment phase; lead time covers from commitment to done. Cycle time (active→done) is a subset of lead time and excludes any queue time between commitment and active work starting.
Slice by — use the dropdown to break FFE down by any column in your data (demand type, team, assignment group). This reveals which types of work sit longest before commitment.
No backlog stages detected. To enable this tab, add your pre-commitment stages (e.g. Idea, Backlog, Ready) to the workflow stage list in Config and drag them above your Start column. Stages before the Start column are treated as the fuzzy front end.
Learn about the commitment point and FFE →
Learn about the commitment point and FFE →
Blocked Items
Impact of blocking on lead time, and which items are currently stuck.
Imported Data
AI Prompt Generator BETA
Generates a structured prompt from your data — copy it into any AI tool. Nothing is sent automatically. You control exactly where it goes — Copilot, Claude, ChatGPT, or anything else.
How it works: Click Generate to build a prompt from your current data. Copy it. Paste it into your preferred AI assistant. The AI has no access to your data — only what you choose to paste.
Data Quality
Every row in your dataset is classified below — so you can see exactly what's feeding your metrics and what isn't. Imperfect data is normal; knowing about it helps you coach better habits.
No data quality issues found.
Monte Carlo Forecast
Monte Carlo simulation runs 10,000 trials by randomly sampling your actual weekly throughput history. Because it uses real delivery patterns — including slow weeks, fast weeks, and everything in between — the ranges reflect genuine variability, not a straight-line average.
Reading the confidence levels — higher confidence means a more conservative (smaller) guarantee. "95% confidence: 37 items" means the team completed at least 37 items in 95 out of 100 simulated futures — very likely, but a cautious number to commit to. "85% confidence: 42 items" means at least 42 items in 85 out of 100 — the recommended planning number. "50% confidence: 49 items" means the team hit that in only half of futures — optimistic, not a commitment. So: the higher the confidence %, the lower the item count. Use 85% as your planning target — realistic without underselling your team or overpromising to stakeholders.
How many by date? — Pick a target date. The simulation tells you how many items you can realistically deliver by then. The backlog list below shows a cut-line at each confidence band so you can see which specific items are inside the commitment.
When will we finish? — Pick an item count (defaults to your current WIP + backlog). The simulation tells you the likely completion date at each confidence level. Higher confidence = later (safer) date. Lower confidence = earlier date but higher risk of missing.
Compare to previous — Import snapshot files exported from earlier runs to track how your forecast has shifted. A later 85% date or fewer 85% items means scope grew or throughput dropped. You can compare snapshots without running a new simulation — just import two or more files directly.
Reading the confidence levels — higher confidence means a more conservative (smaller) guarantee. "95% confidence: 37 items" means the team completed at least 37 items in 95 out of 100 simulated futures — very likely, but a cautious number to commit to. "85% confidence: 42 items" means at least 42 items in 85 out of 100 — the recommended planning number. "50% confidence: 49 items" means the team hit that in only half of futures — optimistic, not a commitment. So: the higher the confidence %, the lower the item count. Use 85% as your planning target — realistic without underselling your team or overpromising to stakeholders.
How many by date? — Pick a target date. The simulation tells you how many items you can realistically deliver by then. The backlog list below shows a cut-line at each confidence band so you can see which specific items are inside the commitment.
When will we finish? — Pick an item count (defaults to your current WIP + backlog). The simulation tells you the likely completion date at each confidence level. Higher confidence = later (safer) date. Lower confidence = earlier date but higher risk of missing.
Compare to previous — Import snapshot files exported from earlier runs to track how your forecast has shifted. A later 85% date or fewer 85% items means scope grew or throughput dropped. You can compare snapshots without running a new simulation — just import two or more files directly.
Probability-based delivery ranges from your historical throughput. Not a commitment — a planning tool.
These are probability ranges, not commitments. They assume throughput stays roughly consistent and scope stays fixed. Both will change. Use the 85th percentile as your planning boundary — it's realistic without being reckless. The 50th percentile will be wrong half the time.
Burn Up
Cumulative completions vs a fixed scope target. Projection lines show when you'll finish at current throughput.
Scope line (blue) — a flat horizontal line showing the total items in scope as of your chosen scope date. Set the scope date to any point in time to fix your target count.
Completed line (green) — running total of items that reached Done each week. Should climb steadily to meet the scope line.
Forecast line (dashed green) — a straight projection from the last completed point, based on average weekly throughput over the last 8 weeks. The date where it meets scope is your expected completion.
±% scenarios — tick "Show scenarios" to draw two extra lines: one assuming you go X% faster, one X% slower. All three projected completion dates are shown in the summary bar.
Scope date — items whose earliest workflow date falls on or before this date count as scope. Default is the latest entry date in your data.
Filters — active filters are respected. Use slice filters to focus on a specific type, priority, or team.
Time in Stage
Median and 85th percentile days spent in each stage, across completed items.
Bars (P50) — the median time items spent in this stage. Half moved through faster, half took longer.
P85 bar — 85% of items left this stage within this many days. A wide gap between P50 and P85 means high variability — investigate why items get stuck there.
Colours — blue = active work stage, amber = queue (waiting). Stage type is also shown in the label. Configure in Column Mapping.
Blocked time row — shown in red at the bottom when blocked data exists. This is the P50 and P85 of total blocked days per item across all completed work. It's shown separately because blocked time is recorded per item, not per stage — we can't tell which stage the block happened in.
What to look for — stages with the highest P50 are your bottlenecks. Queue stages where P85 dwarfs P50 are where work sits unattended. A high blocked P85 relative to stage times means blocking is a significant drag on lead time.
Arrival vs Departure
Items entering the workflow vs items completing each week. When arrivals consistently outpace departures, WIP grows.
Arrivals (orange) — items that entered the active workflow each week.
Departures (blue) — items completed each week (same as throughput).
Net WIP bars — weekly difference (arrivals minus departures). Red = WIP growing that week. Green = WIP shrinking.
What to look for — sustained orange periods signal WIP explosion in progress. This is the early warning sign that cycle times are about to rise.
Built by Simeon Herbert · MIT licence ·