The Workaround Ledger: Putting a Number on the Fixes That Never Got Fixed

A finance team closes the month with a spreadsheet that stitches together two systems which were never designed to speak to one another. A regional manager routes a £4,000 purchase through a colleague's cost centre because the proper approval route takes nine days. A service desk keeps a private list of known errors because the official register is refreshed quarterly.

None of these is a scandal. Each was a sensible response to a constraint at the moment it was made, by someone trying to get the job done. Taken together, however, they form a liability that no balance sheet records and no budget holder owns.

Optimisation debt, defined plainly

Technical debt is the cost of choosing a quick solution now in exchange for slower progress later. Optimisation debt is the same trade applied to how work flows: the accumulating cost of fixes that solved today's problem at the expense of tomorrow's capacity.

It differs from technical debt in three ways. It lives in habit rather than in code, so it disappears from view the moment the person who built it moves on. It compounds, because workarounds tend to grow with transaction volume. And it is fiercely defended, because on the day it was created it was the right answer.

The symptom is rarely dramatic. It shows up as an organisation that can launch one new thing at a time, that needs three weeks to answer a simple question, or that cannot absorb a modest increase in volume without hiring.

Four species of workaround

Spreadsheet sprawl. Critical logic held in files on individual drives: pricing models, allocation tables, compliance trackers. The risk is not the spreadsheet itself but the fact that nobody can say with confidence which version is authoritative.

Shadow approvals and side channels. Informal routes that exist because the formal route is slow, opaque or unpredictable. A quick message to a director gets a purchase moving; the documented workflow sits untouched. The control still appears to operate on paper, which is precisely the problem.

Duplicate reporting. The same figures compiled in two or three places for different audiences, with reconciling effort absorbed invisibly by a middle manager each month.

The permanent temporary. A role, a team or a meeting created to bridge a gap during a restructure, still running four years later with no mandate, no budget line and no review date.

Building the ledger

A workable audit runs over roughly six weeks and does not require new software. The discipline matters more than the tooling.

1. Inventory without judgement. Ask each function for the things they do that are not covered by a documented process. Give a two-week window and a one-page template. A 150-person organisation should expect 30 to 80 entries. Say clearly, in writing, that nothing on the list will be treated as a disciplinary matter.

2. Attach time. For each entry, record who performs it, how often and how long it takes. Sample rather than estimate where possible: ask three people to time the task for a week. Convert everything to annual hours and then to full-time equivalents.

3. Attach exposure. Time is the easy half. Score each item from one to five on four dimensions: key-person dependency, control or regulatory weakness, customer impact if it fails, and data quality risk. A workaround that takes two hours a month but sits on a single person's laptop is not a small item.

4. Estimate the interest rate. Ask whether the cost is fixed or growing. A monthly reconciliation is a steady expense; a manual check that scales with order volume is compounding, and its true cost is the cost it will carry in three years.

5. Rank by repayability, not by size. The largest item is rarely the best first move. Look for entries that are cheap to clear, expensive to keep and unlikely to be defended.

A worked illustration

Consider a 140-person organisation that logs 46 workarounds. The time-attribution exercise finds 210 hours a month devoted to them, or 2,520 hours a year. At a fully loaded cost of roughly £40,000 per productive FTE, that is close to £70,000 of capacity consumed by work that appears in no plan.

The distribution is more useful than the total. In most audits, a minority of items accounts for the majority of hours. Here, 11 of the 46 entries carry around 70 per cent of the time. Three sit with a single named individual and no documented backup. Two represent control gaps that an internal auditor would raise as findings.

That is enough to act on. Two of the largest items can be retired outright, because the requirement that created them no longer exists. Four can be formalised: documented, resourced and given an owner. Three justify modest automation. The remaining items stay on the ledger with review dates.

Turning the audit into a repayment plan

Every entry gets one of three verdicts. Retire it, because the underlying need has gone. Formalise it, because the workaround is genuinely the better design and deserves to become the standard. Or automate it, where the logic is stable and the volume justifies the investment.

Three governance habits make the difference between an audit and a permanent improvement. First, give every workaround an expiry date; a fix without a review date becomes infrastructure by default. Second, ring-fence capacity for repayment, in the region of 5 to 8 per cent of operational effort each quarter, and protect it in the same way as project spend. Third, appoint a single executive owner. Distributed ownership of optimisation debt means no ownership at all.

The objection worth taking seriously

Expect resistance, and expect it to be justified. People defend workarounds because the official process let them down. If a shadow approval route exists, the standard route probably has a defect. The audit should therefore end with a question about the standard, not only about the workaround: why did a capable person conclude that this was the only way to get the work done?

There is also a failure mode in the audit itself. A ledger that grows into a 200-line spreadsheet reviewed quarterly has simply become another piece of optimisation debt. Keep it to one page per finding, one decision per finding, and a firm closing date.

The purpose is not tidiness. It is capacity: the difference between an organisation that can take on the next challenge and one that is already full.