The Second Pass: Measuring the Work Your Business Does Twice
Photo: Francisclarke, CC BY-SA 4.0, via Wikimedia Commons
Most UK SMEs can state what they produced last month. Far fewer can say how much of it had to be produced a second time. Rework rarely announces itself. It appears as a credit note raised after a delivery error, a complaint resolved by a manager on a Friday afternoon, a quote revised because the specification was never fixed, or a team staying late to correct a handover that was never written down. Each event looks like an exception. Taken together, they form a hidden factory: a second operating layer that consumes capacity without appearing on a job sheet or in output figures.
For 2026, the case for measuring it is straightforward. Capacity is expensive, margin is thin, and the work that is done twice is paid for once. A rework audit is not an exercise in blame. It is a process signal, and it is one of the fastest routes to recovered margin available to a lean team.
Why the hidden factory stays hidden
Rework has no natural cost centre. The person correcting the error is usually already on the payroll, the correction is absorbed into the day, and the customer may never be told that something went wrong. Credit notes are filed under finance. Complaints are filed under customer service. Missed deadlines are filed under project management. The connection between them is rarely made, so the same cause is paid for in three different places.
The result is a distorted picture of performance. Output looks acceptable, utilisation looks high, and yet the business feels busier than its sales figures justify. That gap is often the hidden factory at work.
Read the records you already hold
A practical audit starts with evidence the business already produces. Pull the last twelve months of credit notes and refunds. Pull the complaint and query log. Pull the list of jobs, orders or projects that missed their promised date. Then ask a sample of staff to note, for two weeks, the corrections and workarounds they perform that are not on a job sheet.
The aim is not precision to the nearest hour. It is pattern recognition. A rough count of recurring events is more useful than a perfect measurement of a single incident.
Four traces to follow
The first is the credit note trail. Group credit notes by reason, not by customer. Delivery errors, specification mismatches, damaged goods and invoicing mistakes will usually separate into distinct causes.
The second is the complaint log. Read for repetition rather than severity. Three complaints about the same handover are more informative than one dramatic complaint about something that happened once.
The third is deadline slippage. Compare promised dates with actual dates and ask why the gap opened. The answer is often a missing input, an unclear approval or a dependency that was never confirmed.
The fourth is informal time. Ask team leaders where the unscheduled work comes from. The answers will point to the rework that never reaches a report.
Choose three causes, not thirty
A common mistake is to turn the audit into a long list of faults. That list becomes a document nobody owns. A better approach is to score each pattern on three tests: how often it happens, what it costs in time or margin, and whether the business can control the cause. The three patterns with the highest combined score should be the first to fix.
In many UK SMEs, the same three causes recur. Specification is not fixed before work begins, so the job is scoped twice. Handover between sales and delivery is verbal, so the promise made to the customer is not the promise the delivery team receives. Approval sits with one person, so work waits and then gets rushed. These are process issues, not people problems. Treating them as individual failings hides the system that produced them.
Tolerate, reduce or eliminate
Not all rework is waste. A sensible service recovery can protect a long-term account. The test is whether the rework is a deliberate choice or an unmanaged habit. Rare, high-value interventions can be tolerated. Predictable rework should be reduced. Systemic rework should be eliminated at the cause.
Write the decision down. A short register of tolerated exceptions stops the same debate returning every month and makes the cost visible to the people who approve it.
A four-week audit rhythm
Week one is collection. Week two is categorisation. Week three is root-cause discussion with the people closest to the work. Week four is a decision: what will be tolerated, what will be reduced, what will be eliminated, and who owns each action. Repeat the review quarterly. The first audit is always the hardest; the second is where the improvement becomes visible.
The hidden factory will never appear on a job sheet unless someone chooses to look for it. Measuring rework is an act of operational honesty. It replaces the assumption that the business is merely busy with evidence about where capacity is actually going. For an SME with little room for wasted effort, that evidence is not admin. It is margin.
