Imagine a fit-out practice where a restaurant’s joinery specification changes after the latest drawing set has been issued. The change reaches the site team, but the contractor somehow starts work from the old version. By the time Daniel catches the mismatch, the client has already arranged an inspection, and so a replacement drawing must reach the site overnight. The founder, Mara takes the drawing home, works until two in the morning to make the changes, and sends the file before dawn. The client receives the corrected drawings before breakfast. She has protected a real promise to the client and deserves thanks for that night. Everyone in the studio calls it a victory.
The problem becomes clear on Monday. The team discusses Mara’s impressive rescue without examining why an outdated drawing reached the site. Nobody identifies who was responsible for the revision register, the record of changes to the drawings. Nobody checks whether Daniel could have spotted the mismatch before the contractor saw it. The studio celebrates the hero without changing the process that made it necessary.
What the rescue solves, and what remains
A rescue has two outcomes. The immediate outcome is whether harm was contained. Was the client protected? Was the safety risk addressed? Was the deadline met? Those results matter. The second outcome is that the business needs to establish whether the conditions that caused the incident have changed. A rescue can fully protect the client while leaving the business exposed to the same failure happening again.
Repeated rescues can also become part of the firm’s reward system. People who interrupt ordinary work to save the day receive visible praise, while those who quietly prevent a crisis may receive none, because an incident that never happened offers no dramatic story to tell. Employees learn which work earns recognition, and the founder learns to associate being indispensable with leadership. Meanwhile, the process remains vulnerable to another emergency.
None of this makes an emergency an opportunity to demonstrate performance. Mara should not have let the contractor continue simply to demonstrate that the process was flawed. When people, property, contractual promises or a client’s critical deadline are at risk, the first task is to protect them through people who have the appropriate authority to act. Reviewing the incident afterward cannot replace urgent action.
Nor does every rescue reveal a poorly designed process. A storm can close a site, and a supplier can fail without warning. Some tasks genuinely require a qualified professional to intervene. Even a resilient company still needs people who can respond under pressure. The important distinction is whether a similar, foreseeable failure has happened before and whether anyone changes the conditions that made it likely.
Look beyond the late nights
Start with the last few rescues, rather than counting how often people worked late. For each incident, record what first made the rescue necessary, where the problem could have been detected earlier and what the rescuer actually did. Also note whether the rescue involved a decision that had to be made by a person or role with specific authority. Then check whether the cause had appeared before. A one-off storm and the third use of an outdated drawing should not be treated as equivalent examples of heroism.
The appropriate response might be to update the register of issued drawings, clarify a handoff, or establish a way to report bad news before a deadline becomes impossible to meet. In some cases, there may also have been no reasonable way to prevent the incident. Recognizing that limit helps the company avoid introducing controls that cost more than the risks they address.
Prevention is harder to assess when someone receives no credit for preventing an incident. Suppose Priya notices a revision mismatch on Tuesday and asks the team to hold back the drawing set until the mismatch has been checked. A project manager may later remember Priya as the person who delayed the drawings, while remembering Mara’s overnight rescue as the act that protected the deadline. Neither description tells the whole story. Priya’s intervention may have been more valuable than Mara’s rescue, or it may have been excessive caution. The studio needs evidence to distinguish those possibilities.
A proper incident review should therefore record both what the rescue achieved and what it consumed. Record the hours diverted from other work, any risk introduced by a rushed repair, the dependence on one person, and the changes actually made afterward. Fatigue is only part of the cost. If every important exception waits for the company’s most skilled rescuer, the next crisis may arrive while that person is absent. The business then depends on the rescuer’s availability, rather than a defined process that people are authorized to carry out.
The question for the Monday meeting is not whether Mara deserves praise. She does. It is whether the studio can also explain what it will do differently before the next emergency, now while nobody is standing at a drawing board after midnight.



