Time Metrics: Throughput, Processing, and Waiting

Time Metrics

Time shows up in almost every chart, and the numbers can look very different from one place to the next. That is not a mistake: ProcessMind measures time in three ways, and each way answers a different question. This page explains what each measure counts, so you can trust the number in front of you — and pick the right one for the question you are asking.

The three time measures

Measure	What it counts	Counted per
Throughput time	How long a case takes from start to end — or between the From and To elements when the chart limits the span	case
Processing time	Time actually spent working on a step, from the moment work starts until the step is recorded	step
Waiting time	Time a step sits idle, from the previous step in the case until work starts on it	step

The rule behind everything else on this page: throughput time describes cases, processing and waiting time describe the steps inside them.

Throughput time — the case

• It measures the elapsed time of the case. If the chart has a From element and a To element selected, it measures the span between those two points instead — for example from “Order received” to “Invoice sent”.
• It is counted once per case. In trend charts a case is placed in the time bucket where it ends, so a case is never counted twice.
• Only closed cases contribute by default; an open case has no end yet. See Closed and open cases.
• The Sum version adds all cases together, so it grows with the number of cases — not with the number of events.

Processing time — active work

• It measures work done at a step, for example how long an approver actually spends on an invoice.
• It is counted per step occurrence, so a case that repeats the same activity three times contributes three values.
• Use it to find activities where the work itself is heavy, not where the work is waiting.

Waiting time — idle time before a step

• It measures the gap between the previous step in the case and the start of the next one.
• It is counted per step occurrence, just like processing time.
• Use it to find queues, batching, and handoffs — the places where a case is not being worked on but is not moving either.

Reading the three together

A case can take 12 days from start to end while the work on it adds up to a few hours. The rest is waiting time: queueing, approvals waiting in an inbox, batched processing, and handoffs between people or systems.

Two numbers on a dashboard summarize that split:

• Active step share — the part of measured step time spent working instead of waiting. A low value tells you the visible slice is mostly idle between touches.
• Slow-case tail — P90 cycle time compared with the median. A high multiple means a small group of cases is stuck far beyond the typical case.

The wording you see

The same measure is named differently depending on where you look, because each place answers a different question.

Wording	What it means	Where you see it
Throughput time	The measure itself, for cases	Metric lists, chart labels and the standard KPI tiles: “Average throughput time”, “Median throughput time”, “Avg. case throughput time”
Cycle time	The same case-level number, in plain language	Dashboard charts and tiles: “Average cycle time per vendor”, “Avg. cycle time”, “Median cycle time”, “P90 cycle time”, “Slow-case cycle time over time”
Total cycle time	All cases added together	The Sum option in the metric list
Transition time	Time between two connected activities on the model, counted per transition — not per case	The connection menu and details panel: “Avg. transition time”
Case duration	The case-level measure when you slice instead of measure	The Case duration filter and the “Slow cases” drill-down
Waiting time and Processing time	The two step-level measures	Their own metrics, for example “Average waiting time per activity”

Duration Metrics in Custom Dashboards

Custom dashboards let you configure duration metrics for the question you want to answer. Select the appropriate duration measure, such as throughput, processing, or waiting time, and configure the chart to show the relevant aggregation for your analysis.

A few rules of thumb:

• Metric lists and axis labels use throughput time.
• Dashboard titles and tiles often use cycle time for the same number.
• On the connections of the model the number is transition time, because it describes one transition between two activities.
• Filters use case duration.
• Waiting and processing always refer to step-level time — never to the case.

Why a chart can show days while a connection shows hours

Both numbers are correct. The chart measures the whole case, so it shows days. The connection measures the time between those two activities only, so it shows hours. Use the chart to answer “how long do our cases take”, and the connection to answer “how much time passes between step A and step B”.

Closed and open cases

Throughput time needs an end. A case that is still open has no end yet, so it cannot contribute a throughput time — that is why charts of throughput time use closed cases by default.

• Change it per chart: open the chart settings and set Cases to All, Open, or Closed.
• Waiting time and processing time are measured on the steps themselves, so they do not require the case to be finished.

Next step

Open a dashboard and check two charts next to each other: one showing case numbers over time, and one showing cycle time. Then use Metrics to switch a chart to waiting or processing time and see where the time in your process actually goes.

• Metrics: the full list of metrics and what each one shows.
• Process Charts: the chart types that display these metrics.
• Using Filters: scope charts to the cases you care about.
• Process Graph: read transition time and other metrics on the connections.