The Rule Nobody Could Remove
A small mistake, a permanent form
Imagine a business that once sent an order to the wrong branch. It introduced a second address check. Later, somebody added a confirmation email. After another misunderstanding, a spreadsheet appeared. Years passed. The ordering system improved, the customer details became easier to verify, and the original workaround stayed. A routine order now travels through several reminders of a problem that nobody has recently examined.
This is a hypothetical business, not a reported case. Its rules need not be foolish. Each may have been a reasonable response to a particular difficulty. The trouble begins when the business preserves every response but stops asking whether the difficulty still exists, whether another control already covers it, or whether the extra step creates a new source of error.
Adding a procedure has a visible author and a reassuring explanation. Removing one can look like withdrawing care. That asymmetry makes accumulation easy: the business remembers who introduced the protection, but may have nobody responsible for checking whether it still protects anything.
An addition is easier to show
A manager asked to improve a process can produce a new checklist, meeting or report. There is something to present. A removed duplicate field leaves no impressive document behind. Its benefit appears in work that no longer needs doing, often spread across people whose individual saving seems too small to mention.
Research by Adams and colleagues suggests that people can overlook useful subtractive changes. That is a prompt to consider removal, not evidence that fewer rules always make a better business. The practical question is whether an addition is being chosen because it solves the problem, or because it is the first kind of solution the team thinks to propose.
In the imagined order process, the answer to a confusing spreadsheet might be a clearer spreadsheet. It might also be to stop maintaining a second record when the agreed order system already contains the information. Those are different proposals. Both deserve consideration before somebody schedules training on the latest version of a document that may not need to exist.
Find the job the rule is doing
Before removing the second address check, ask what it actually catches. Follow a few recent orders through the real process, including an exception. Who checks which address, against what source, and what happens when the two records disagree? A tick beside a field does not establish that anyone compared the right information.
Speak to the people who use the result, not only the person who completes the step. A weekly report that looks unread may feed a later decision. A copied detail may be compensating for a system limitation that managers never see. Removing it without understanding that dependency can simply move the work to somebody less visible.
Ask also who owns the rule and what authority is needed to change it. Legal, contractual, security and safety obligations require appropriate review. An absence of recent incidents does not prove a safeguard is useless; it may be doing its job. If the purpose is unclear, investigate it. Do not turn uncertainty into permission to delete.
Separate the purpose from the ritual
Suppose the review finds that the customer confirms the delivery branch in the current order system, while staff copy the same address into a separate sheet solely to show that it has been checked. Nobody uses the sheet to make another decision. That is a candidate for simplification, provided the remaining record is reliable and the necessary check still happens.
The purpose is a correct delivery destination. The ritual is one particular way of documenting attention. Keeping the purpose does not necessarily require keeping every historical action around it. The team might remove the duplicate entry, improve the source record, or restrict an additional check to cases where the address has actually changed.
Do not assume that automation settles the question. Automatically copying an unnecessary record can make duplication faster while preserving disagreements and maintenance work. First decide which information should be authoritative, which check is needed and who must act on a mismatch. Only then decide whether software should perform part of that job.
Remove one thing, watch the whole job
For a low-risk internal step that the responsible owner is authorised to change, agree a limited trial. Specify what will stop, what protection remains, which orders are included and how staff can restore the old process if a defined problem appears. Brief everyone affected. A rule that disappears from one person's screen but survives in another person's instructions has not been retired.
Look beyond the time saved at the removed step. Check total handling time, incorrect deliveries, rework, missing information and questions passed to other colleagues. Ask whether the customer now has to supply something the team previously handled. A shorter internal process is not a better service if it exports the work to the buyer.
A quiet fortnight is useful operational evidence, but it cannot establish that rare failures will never occur. Consider the volume and variety of cases observed and retain appropriate controls. If the trial exposes an important dependency, restore or redesign the step. Reversal is part of a controlled trial, not an embarrassment that must be hidden to protect the idea.
Do not make deletion the next target
A target to remove ten rules can reproduce the problem it claims to solve. Teams may delete easy wording, merge documents without reducing work, or discard a useful check to reach the number. Count a rule removed only when the affected work has actually changed; judge the result by what the process needs to accomplish.
When a new temporary procedure is introduced, give it an owner, a reason and a review point. The review need not mean automatic expiry. It means someone will ask whether the original condition persists and what evidence supports keeping, changing or ending the procedure. Keep that record proportionate. A committee for retiring every small checklist can become the next checklist worth questioning.
Return to the business with the misdirected order. The goal is not to forget the mistake. It is to keep what was learned without making every future customer pay for every form created afterwards. A well-run business can explain why a step exists. It should also be able to recognise when that explanation no longer holds.
Further reading: Adams and colleagues (2021). The business scenario and practical proposals are editorial analysis, not findings from that study. Adams et al. (2021).
Leave a comment