Process Debt: The Workarounds Quietly Costing You Margin

What is process debt?

Process debt is the accumulated cost of every workaround, exception and verbal agreement your team relies on because the underlying process was never fixed. It is not disorder and it is not incompetence. It is interest, and you are paying it on a decision somebody deferred, usually for a sensible reason, usually in a busy month.

The working definition we use inside client engagements: process debt is the gap between how your business is supposed to run and how it actually runs, held open by people quietly doing extra work.

The borrowing is real, and engineering teams worked this out first. In a survey of more than 1,000 developers and more than 1,000 C level executives across the United States, the United Kingdom, France, Germany and Singapore, developers reported spending 17.3 hours a week addressing technical debt (Stripe, The Developer Coefficient, September 2018).

What engineering did next is the part operations teams have not copied. They gave the debt a name, put it on a board next to the new work, and paid it down on a schedule. Operations debt gets none of that treatment. There is no register, no owner and no repayment plan, so it compounds quietly and it is never anybody's job this week.

The distinction that matters: a bad process is a design problem you can see. Process debt is the invisible layer of human effort holding a bad process together. You fix the first with a redesign. You clear the second by finding out who is paying the interest.

Why does process debt never show up on the profit and loss?

Because every item of it is individually too small to fight for, and collectively large enough to move your margin. Three mechanics keep it off the page.

It is paid in minutes, not invoices. Slack's Workforce Lab surveyed 10,281 desk workers across Australia, the United States, France, Germany, Japan and the United Kingdom and found they report spending 41% of their time at work on tasks that are low value, repetitive or lack meaningful contribution to their core job functions (Slack, Workforce Lab, February 2024). Not all of that is process debt. None of it is billed anywhere, so none of it is managed.

It is paid by your most capable people. A workaround is evidence that somebody is coping. The person who keeps the spreadsheet reconciling two systems is solving your problem, which is precisely why the problem stops being visible. Praise the coping and you have bought the debt permanently.

It transfers rather than resolving. New hires learn the workaround as the process, because that is what they are shown in week one. Tools get configured around the exception. Within two years the workaround has a name, a folder and a defender, and removing it looks like a reorganisation rather than a repair.

Underneath all three sits one structural reason. Almost every workaround lives between two owners, which means it lives with neither. Work inside a function has a manager. Work in the seam has a volunteer, and volunteers do not raise budget requests.

How do you find your process debt? The Process Debt Register

You find it by asking where your team does extra work, then classifying each answer by what is actually missing. There are only four possible answers, and that is what makes the register a diagnostic rather than a list.

A workaround is a person filling a gap. The gap is in the writing, in the coverage, in the ownership, or in the tooling. There is no fifth place for it to hide, so four classes are complete by construction. Every workaround you can name belongs to exactly one of them, and the class tells you what repayment costs.

Exhibit 1: The Process Debt Register

Class	What the workaround replaces	Who pays the interest	What repayment looks like
Undocumented debt	A written process. It lives in one or two heads instead.	The people who hold it, in interruptions. The business, on the day they leave.	One page per process: steps, owner, tools, edge cases. A named keeper who updates it.
Exception debt	Coverage. The process exists, but a known set of cases bypasses it.	The team, in judgement calls that should not require judgement.	Fold the top two exceptions into the standard route, or make them a second route with its own owner.
Handoff debt	An owner for the seam between two functions.	The receiving team, in rework and in chasing.	One named owner for the boundary, one definition of ready, one queue everybody can see.
Tooling debt	Alignment between the system of record and the process.	Everybody, in retyping and reconciliation.	Change the tool or change the process. Keeping both and bridging by hand is the debt.

Running it takes about an hour. Take the three processes that generate the most questions, list every workaround anybody can name, and put a class and an owner against each line. Then rank the register by interest paid per week rather than by how annoying each item feels. Those two orderings are rarely the same, and the annoying items are usually the cheap ones.

Diagnosis is rarely the hard part. Funding the repayment is. In the Singapore Business Federation's National Business Survey 2023/2024, the sixteenth annual edition, fielded from 14 September to 12 November 2023 with 1,056 respondents of which 82% were SMEs, only 25% planned to increase investment in reengineering their business and operational processes over the next year, while 70% planned to keep that investment the same. The same survey found 29% naming a lack of management expertise to drive technological change as a challenge in technology adoption (Singapore Business Federation, National Business Survey 2023/2024).

The prize, meanwhile, is well measured. Asana's Anatomy of Work Global Index 2023, a survey of 9,615 knowledge workers across Australia, France, Germany, Japan, the United Kingdom and the United States, found that work about work, meaning time spent on coordination rather than the skilled work people were hired to do, still accounts for 58% of the day (Asana, Anatomy of Work Global Index 2023). A register does not recover all of that. It recovers the part your own business built.

What does process debt cost in a business this size?

The figures below are illustrative, using the assumptions stated. The company is a composite profile, not a Rem.Up client.

An Auckland based B2B services firm: 62 people, USD 11M revenue, USD 1.32M operating profit at a 12% margin. Thirty eight of the 62 work inside the chain that runs from signed order to delivered work, which is where a register would find most of the debt.

Assumptions, and they are assumptions: a fully loaded cost of USD 300 per person day inside that chain and USD 420 for the two senior people who hold the undocumented process; 46 working weeks; an eight hour day; and only two thirds of the total treated as recoverable, because a share of coordination in any handoff is irreducible.

Undocumented debt. Two senior people answer some version of how do we handle this one again for about four hours a week each. That is 8 hours a week, 368 hours a year, 46 person days, at USD 420 a day: USD 19,320.

Exception debt. The 38 people in the chain each spend about 45 minutes a week on the bypass route for cases the standard process does not cover. That is 28.5 hours a week, 1,311 hours a year, 164 person days, at USD 300 a day: USD 49,200.

Handoff debt. One job a week comes back across the seam between sales and delivery and costs two person days to redo. That is 92 person days a year, at USD 300 a day: USD 27,600.

Tooling debt. Named, not priced. Two systems disagree and a spreadsheet reconciles them nightly. We could price the retyping, but the real cost is the decisions taken on the wrong number, and we cannot price that honestly. It stays on the register with no figure against it, which is more useful than a figure we invented.

Three lines priced: USD 96,120 a year. Two thirds recoverable: USD 64,080 a year, or USD 5,340 a month. That is 4.9% of operating profit.

Notice that 4.9% is not a number that wins a board meeting, and that is exactly why the debt survives. Each line is defensible on its own. None of them is urgent. Notice also how conservative the model is: at roughly 1.5 hours per person a week it prices about a twentieth of the low value time Slack's respondents describe, because most of that 41% is not process debt at all.

Then the debt metaphor earns itself. Nothing in those three lines shrinks on its own, and every line scales with headcount. Grow the team 15% a year with no repayment and the same debt costs USD 73,692 in year two and USD 84,746 in year three. Across a three year planning cycle that is USD 222,518, none of which ever appears as a line item anybody can cut.

Two sensitivities, because a model you cannot stress is a sales number. Halve every assumption and it still reads about USD 2,670 a month. Double the hours per person, which is still a long way below what the survey data would support, and it reads 9.7% of operating profit.

What do you do on Monday?

Run the register once, price one line, and give that line an owner and a date. Three moves, in that order, and none of them needs a budget.

Ask the question that finds the debt. In your next team meeting, ask what people worked around last week. Not what went wrong, which gets you incidents. What they routed around, which gets you the register. Write down every answer and put a class against each one.

Price one line, not four. Pick the line with the most people paying interest, count the hours honestly for two weeks, and multiply. One priced line moves the conversation from tidiness to margin, and a priced line is the only version of this that survives a budget meeting.

Repay on a schedule, one item at a time. One workaround retired per month, with a named owner and a date, beats a process improvement programme nobody staffs. Put the register in the weekly operating review so the balance stays visible, and reprice the top line each quarter.

The test is not whether your processes are elegant. Nobody is paid for elegance. The test is whether the number on the register is going down.

Rem.Up places senior independent operators inside founder led startups and SMEs to run operations as a Fractional COO or to own a critical project as a Fractional PM, across APAC and Australia. If you want a second pair of eyes on your register, thirty minutes is enough to start: we walk your three noisiest processes, classify what we find, and tell you which line to price first. More at rem-up.com, or book a call at calendly.com/john-rem-up/30min.

Innovate. Optimize. Grow.