Every step has an owner. The waiting between them has none.

Draw any process and you get boxes and arrows. Assign responsibility for it and you will find, almost without exception, that every box has a name against it and no arrow does.

That is a problem, because the arrows are where the time is.

The arithmetic of unowned time

The boxes are where work happens and they are usually a small fraction of the elapsed time. Following one expense claim end to end put twenty minutes of work inside a hundred and twenty nine hours of elapsed time, and every one of those twenty minutes had a department attached to it. The hundred and twenty eight hours had nobody.

---

	Time inside the steps, with a named owner	20 min
	Time in the gaps, with no owner	129 h
	Share of elapsed time somebody is accountable for	0.26%
	Share of elapsed time nobody is accountable for	99.74%

Illustrative figures. Every person in this process is meeting their own target. The process is failing, and there is no meeting at which that contradiction is anybody's agenda item.

This is why process reviews so often end with everyone able to demonstrate that their part is fine. Each part genuinely is fine. The thing that is failing is not any of the parts.

Why it happens, and it is not negligence

Ownership follows organizational structure, and processes cross it. A department owns its own work because that is what a department is. The handover between two departments belongs to neither, and there is no structure at that level for it to belong to.

The gap is not visible from either side. From inside a step, work arrives and gets done. The queue in front of it is somebody else’s output rate and the queue behind is somebody else’s input. Neither team sees the elapsed time from end to end because neither team has that view.

And it is nobody’s failure to report. A step that takes three minutes reports three minutes. Nothing in the system generates a number called “waited fifty two hours between finance and payment”, so nothing gets escalated. The time is invisible rather than tolerated.

The three fixes, cheapest first

Measure the gaps and put them on the same report as the steps. This alone changes conversations, because it moves the gap from invisible to merely unowned, and an unowned number that is on a report tends to acquire an owner. It costs nothing but a timestamp at each handover.

Give the end to end time one name. Not the steps: the whole thing, from arrival to completion, belonging to one person whose measure is the elapsed time and who has no ability to make any individual step faster. That constraint is deliberate. Their job is the gaps, because that is the only part nobody else is looking at.

And attach the gap to the receiving step rather than to nobody. If finance owns “claims waiting to be checked” as well as “claims being checked”, the queue appears on somebody’s report the moment it grows. This is the cheapest structural change available and it is usually resisted, because it makes a team accountable for a number partly driven by somebody else’s release pattern. That objection is correct and it is still better than the alternative, which is nobody owning it at all.

What to do with it

Take your process map and put a name against every arrow. Not the boxes, the arrows. The exercise takes ten minutes and the result is usually a row of blanks, which is the finding.

Then measure the three biggest blanks for a fortnight. Two timestamps per handover. You do not need a system; a spreadsheet is enough to establish whether the gap is hours or days.

And check whether the gap is queued, batched or scheduled, because those have completely different fixes. A queue is a capacity question. A batch is a policy somebody chose. A schedule is a calendar somebody set. Only the first one is about how hard anybody is working, and it is usually the smallest of the three.

The boxes are where the work is. The arrows are where the time is, and almost nowhere are they anybody’s job.