FlowLens

What is cycle time?Cycle time is how long it takes to complete an item, measured from when it enters your start stage to when it enters your end stage (as set in column mapping). It is the customer's experience of wait time.

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 cycle 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.

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 cycle time.

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.

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.

Cycle time (p50) — median time from commitment to done. Lower is better. Rising cycle 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 cycle 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 →

Little's LawThe most important equation in flow metrics: Cycle Time = WIP ÷ Throughput If WIP doubles and throughput stays flat, cycle time doubles. You cannot improve cycle time without managing WIP.

Rising WIP means more is being started than finished. This is the single most common cause of long cycle 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.

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.

What this showsThe shape of your delivery. Each bar = how many items completed at that cycle 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 cycle times — consider whether they should be tracked separately.

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 cycle 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.

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 cycle 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.

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 (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 an item moves from the backlog into active flow. Cycle time should always be measured from this point, not from creation. Items can sit in the backlog for months — including that time inflates cycle time and hides delivery capability.

Lead time = FFE time + cycle time. This is the total elapsed time from backlog entry to completion — what the business or customer actually experiences.

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.