Loading page
About this tool
Find slow workflow steps by comparing active time, wait time, rework, and throughput pressure.
The Workflow Bottleneck Finder splits each step of a process into active minutes, wait minutes and rework minutes, adds them into a total per step, and names the step with the largest total as the bottleneck. Paste one line per step as name, active, wait, rework and you get the end-to-end cycle time, a bar for every step scaled against the worst one, and the share of each step that is pure waiting. It is the standard value-stream split of touch time versus queue time, done without a spreadsheet.
Open Workflow Bottleneck Finder on AltFTool — it loads instantly in your browser.
Paste one step per line into the "Step, active minutes, wait minutes, rework minutes" box, like Approval,10,96,8 — a five-step sample workflow is preloaded.
Each line's three numbers are totalled live, and every bar in the Bottleneck map is scaled against the largest step total with its wait-share percentage.
Read the Cycle time, Bottleneck and Steps cards to see the end-to-end minutes and which step carries the most drag.
Every step reports a wait share as a percentage of its own total, which is what distinguishes a genuinely hard step from one stuck in a queue.
Time spent redoing a step is tracked separately instead of being buried in active time, so quality problems show up as time problems.
Each bar is scaled to the largest step total, so the size of the gap between your bottleneck and everything else is visible at a glance.
One step per line as name, active minutes, wait minutes, rework minutes — for example "Approval,10,96,8". All four fields are required; lines missing a field are ignored, and any non-numeric value is read as 0.
It is the step with the highest total of active plus wait plus rework minutes. That means a step can be flagged as the bottleneck on queue time alone — in the sample, Approval carries 96 wait minutes against 10 active, a wait share of 84%.
Use typical or median minutes per step for a first pass, since averages get skewed by rare escalations. If you want to understand missed deadlines rather than everyday flow, run it a second time with the slow-case numbers and compare which step moves most.
No — the total is the straight sum of every line, so it models a strictly sequential process. If two steps genuinely run at the same time, enter them as a single combined line using the longer of the two, otherwise the cycle time will overstate the real elapsed duration.