Every organisation knows which process everybody complains about. It is rarely the one costing the most money. The expensive one is usually quiet, runs many times a day, and nobody mentions it because it has always been like that.
You cannot tell them apart by asking. You can by counting.
Three numbers, and only three
How often does it run? Per week or per month. From the system that records it, not from somebody's memory. If nothing records it, that is the first finding.
How long does it take, in people's time? Not elapsed time. A request that sits for three days and takes eleven minutes of human attention costs eleven minutes. Elapsed time matters to whoever is waiting; only attention costs money.
What is that time worth? Salary plus employer costs, divided by working hours. A rough figure is fine, and being open about how rough it is matters more than the number: a defensible estimate everybody can see beats a precise one nobody can check.
Multiply the three. That is the annual cost of running it as it is.
Then two more, which change the answer
How often does it go wrong, and what does putting it right cost? Rework is usually where the money is. A process that takes four minutes and needs redoing a third of the time costs far more than one taking six minutes that always works.
What does the waiting cost? Sometimes nothing. Sometimes it is the whole thing: an engineer idle, a shop unable to trade, a customer who goes elsewhere. Count it only where you can name the consequence, and say so when you cannot.
Worked example
A request that runs 40 times a week, takes 25 minutes of attention, at £35 an hour including employer costs:
- 40 × 25 minutes = 16.7 hours a week
- 16.7 × £35 = £584 a week, about £30,000 a year
Now add that one in five needs redoing, at another 25 minutes: £36,000 a year.
That is a number somebody can act on. "Everyone finds it frustrating" is not.
Measure before you change anything
This is the part that gets skipped and it is the part that pays. If you do not measure first, three things happen. You optimise the loud process instead of the expensive one. You cannot prove the change helped. And you cannot tell the difference between a change that worked and a quiet month.
A week of counting before a month of changing is not a delay. It is what makes the month worth doing.
Take steps out before adding anything
The instinct is to automate. Automating a bad process gives you a faster bad process, and now it is harder to change because it is in code.
In order:
1. Stop doing the steps that serve nothing. Approvals that are never refused. Forms nobody reads. Checks that duplicate a later check. 2. Move the decision to the person with the information. Most delay is waiting for somebody who knows less than the person who asked. 3. Then automate what is left, if it is still worth it. Often it is not, which is a good outcome and costs nothing to find out.
When the answer is to leave it alone
Sometimes the honest finding is that a process costs £4,000 a year and changing it would cost £30,000. Then the recommendation is to change nothing, and say so plainly.
That is a real answer and it is worth paying for. A review that recommends work in every case was not a review.
This is part of what we do under Process optimisation. See the services, or tell us what is not working.