top of page

When the SOP stops matching the work

Sep 29
4 min read

A new colleague opens the knowledge base, follows the approved steps, and reaches a dead end. The button in step four has moved. The team responsible for step six no longer exists. A manager has changed the escalation rule, but the document still describes the old one.


Nobody meant to create a bad procedure. The work simply moved, and the writing stayed still.


That gap is one reason I became interested in SOPs. During four years in remote technology teams, I learned that being good at the job rarely meant remembering every answer. It meant knowing where to find reliable information. I would tell teammates that finding the right information mattered more than trying to hold it all in your head. But when teams changed, processes shifted and knowledge lived across documents, systems and people's memories, even a good search could lead to the wrong answer.


The quiet cost of a confident wrong answer

An outdated SOP is more dangerous than an obviously missing one. When there is no guidance, a colleague knows to ask. When a polished document carries an old instruction, they may follow it with confidence. The result can be a delayed case, an unnecessary handoff, a different answer to the same customer question, or a new teammate learning a workaround that should never have become normal.


The effects are cumulative. One person spots a mismatch and asks in chat. Someone replies with the current rule. The next person finds the original document and asks again. Eventually, the unofficial answer becomes more useful than the official one. The team now has two operating systems: what the knowledge base says and what experienced colleagues know.


This is process drift. It is especially common when a company grows, changes tools, reorganises a team, adds a product, or alters an approval rule. A procedure that was accurate when written can become unreliable without anyone formally deciding to retire it.


A document is only one part of the process

Atul Gawande's The Checklist Manifesto argues that simple prompts help people handle work that has become too complex to execute consistently from memory alone. The World Health Organization's surgical checklist provides a striking example: in eight pilot hospitals, its introduction was associated with substantial reductions in major complications and deaths.


A support escalation is not surgery, and those outcomes should not be borrowed as a promise for office teams. The transferable lesson is more modest: clear, usable prompts can help capable people remember and coordinate important steps.


Yet a checklist only helps when it fits the real setting. The Institute for Healthcare Improvement describes sustained improvement as work that requires standard processes and attention to how they are used at the point of care. In an office, that means watching what actually happens after a colleague opens the SOP.


Can they find the instruction? Does it answer the decision in front of them?

Is the escalation route still active? Does the process allow for an unusual customer situation?


An SOP should guide judgement, including when to stop and ask, rather than pretend every case is identical.


Five signals that an SOP needs attention

People keep asking the same question. A repeated question may indicate unclear wording, poor search terms or missing guidance. Do not assume the team simply needs to read harder.


The team has a chat-based correction. If the current instruction is pinned in a channel while the official document says something else, the fix belongs in the source document.


New starters learn the exceptions verbally. When onboarding depends on knowing which pages to ignore, experienced colleagues become an invisible layer of infrastructure.


A tool or owner has changed. A new form, queue, product setting or manager can invalidate a sequence even when most of the page still looks right.


Customers receive different answers. Inconsistent outcomes are a reason to inspect the underlying rule and its communication, not just the individual response.

These signals do not always call for a total rewrite. Often, a short, targeted correction will restore confidence faster than a grand documentation project.


Make change visible to the document owner

A practical review loop can be small. Name an owner for each important SOP. Put a last-reviewed date on the page. Give colleagues a simple way to flag a broken step while they are working. When a policy, system or team changes, list the affected procedures as part of the change itself. Review high-impact instructions more often than low-risk reference pages.


The owner then checks the report with someone who performs the work, updates the source, records what changed, and tells affected teammates. A change log need not be elaborate; a date, a short description and the decision behind it are often enough. The aim is to prevent yesterday's answer from quietly becoming today's training material.


Harvard Business Review has argued that process design benefits from frontline input. That matches my experience: the person following a procedure sees the confusing branch, the extra click and the customer question that a writer at a distance can miss. Invite that person into the review, and make it safe to say, “This page no longer matches what we do.”


Start with one real journey

Choose a frequent task, such as an onboarding request, a billing query or a support escalation. Ask a colleague to complete it using only the current SOP. Note where they hesitate, leave the page, ask another person or apply an undocumented exception. Then update that one procedure and test it again with someone else.


The point is not to make a perfect library overnight. It is to restore the relationship between the written process and the living work. When a team can trust that relationship, people spend less time checking whether guidance is true and more time helping the person in front of them.


Further reading and source notes

 
 
 

Comments


bottom of page