Ampersand Consulting All articles
Operations

When the Workaround Becomes the Work: Closing the Gap Between Process and Practice

Ampersand Consulting
When the Workaround Becomes the Work: Closing the Gap Between Process and Practice

Photo: Vaper City, CC BY-SA 4.0, via Wikimedia Commons

Every organisation has two operating models. One is written down: the process map on the intranet, the workflow in the enterprise system, the procedure the auditor was shown. The other is what people actually do on a Tuesday afternoon when a customer is waiting and the system will not cooperate.

The distance between them is usually small at first. Then it compounds. A few years later, the documented process is a work of fiction, the real process lives in the heads of four long-serving employees, and nobody can say with confidence what the organisation does when one of them is on leave.

This is exception creep, and it is one of the least examined sources of margin leakage in UK operations.

The gap you never authorised

No one decides to build an informal operating model. Exceptions arrive one at a time, each with a reasonable justification. A key account needs a bespoke invoice format. A supplier cannot use the portal. A regulatory deadline falls on a weekend, so a colleague authorises a manual release. A defect requires a same-day fix, so an engineer bypasses the change board.

Each decision is locally sensible. The problem is that none of them is reviewed in aggregate. Individually they are exceptions; collectively they become the operating model, and the operating model is never designed, costed or approved by anyone.

How it accumulates

Four mechanisms do most of the damage.

Local optimisation. Teams optimise for their own service level, not the organisation's. A workaround that saves the service desk twenty minutes may cost the finance team two hours of reconciliation, but nobody is measuring the second figure.

System friction. Where a system is slow, inflexible or simply disliked, people route around it. Workarounds are rational responses to bad tools, which is why telling people to stop using them rarely works on its own.

Customer accommodation. Bespoke handling is often the price of a large account. It is rarely priced into the contract, and it rarely stops expanding once the precedent exists.

Workaround inheritance. New starters learn the job from the people beside them, not from the documented process. Within two years, the workaround is the training, and the original method has quietly disappeared. This is the most corrosive mechanism, because it converts a temporary fix into institutional knowledge.

The bill arrives in three places

Margin. Exceptions consume time that is rarely billed and almost never measured: rework, expediting, manual reconciliation, small-batch handling, the second approval that exists only because the first is not trusted. In professional services this shows up as falling realisation rates; in manufacturing and distribution, as changeover time and stock anomalies.

Risk. Informal paths lack audit trails. When a complaint escalates, an insurer asks questions, or a regulator requests records of how a decision was reached, the organisation has to reconstruct its own behaviour from emails. For firms operating under FCA rules, Consumer Duty expectations or ISO-certified quality systems, unrecorded deviation is not merely untidy — it is exposure.

Staff fatigue. Exceptions live in working memory. The person who knows that client X needs the alternative form, that system Y rejects certain postcodes, and that Z must be manually checked carries a cognitive load that no job description acknowledges. That load is a leading indicator of resignation, and it is concentrated in exactly the experienced people an operation can least afford to lose.

Signals worth looking for

Exception creep is visible before it is measured. Watch for the same deviation occurring more than twice a month; for approvals that only one named individual can give; for spreadsheets that sit alongside the main system rather than inside it; for onboarding times that lengthen despite stable headcount; for credit notes and rework rates that drift upward; and for sentences beginning "we just know to…".

A short audit that produces evidence

Pick a sample of twenty to thirty recent transactions — orders, claims, invoices, cases, jobs. For each, record the documented path, the path actually taken, who intervened, why, and how much time the deviation added. Tally the results.

The exercise typically takes a week of one person's time and produces something an organisation rarely has: a quantified description of its real operating model. Repeat quarterly until the pattern is under control, then annually.

Deciding, not listing

An exception register that nobody acts on is just another form of documentation drift. Sort each finding into one of three responses.

Formalise. The exception reveals a genuine need the process does not meet. Build it into the process properly, with a system field, a documented rule and a price where appropriate. Formalising is not defeat; it is honesty.

Automate. Many exceptions exist because of a fixable friction — a missing data field, an over-restrictive permission, an approval that requires a signature rather than a click. Removing the friction removes the workaround.

Eliminate. Some accommodations are historical rather than necessary. Withdrawing them requires a conversation with a customer or a policy line from a director, and it is usually deferred for that reason. Deferral has a cost, and the audit gives you the number to put beside it.

Where an exception is politically difficult, escalate it as a decision rather than a task. The purpose of the exercise is to force a choice, not to build a longer list.

Keeping it closed

Exceptions reopen. A rule that is not monitored decays within two quarters. Give the register an owner, set a threshold above which no manager may grant an exception unilaterally, and review the count in the monthly operations meeting alongside service levels and cost.

Above all, treat practice as evidence. If the documented process and the real process disagree, one of them is wrong. The organisation's job is to decide which — deliberately, and in writing — rather than letting the answer be settled by whoever is most tired.

All Articles

Related Articles

The Invisible Wiring: Mapping the Dependencies Your Org Chart Misses

The Invisible Wiring: Mapping the Dependencies Your Org Chart Misses

Finding the Gaps Between Teams: A 90-Minute Handoff Audit for UK Operations

Finding the Gaps Between Teams: A 90-Minute Handoff Audit for UK Operations

The Discipline of Not Starting: A Five-Question Screen for New Initiatives

The Discipline of Not Starting: A Five-Question Screen for New Initiatives