How to Escalate Inside an Institution Without Burning It Down

Why this matters

You have found something the building needs money for, and your contact cannot approve it. Every shop's instinct at that point is to find someone who can, which usually means going over the contact's head. That move works about as often as it ends the relationship, and when it ends the relationship it ends it permanently, because you have publicly demonstrated that your contact does not control their own building.

The version that works does not move up. It moves sideways, into whichever of the institution's own risk categories your finding actually belongs to, and it arrives with your contact holding it rather than being bypassed by it. An institution will spend money it does not have on a safety exposure and will not spend money it does have on a vendor's recommendation. Same finding, same building, different routing.

The six steps below are ordered by what skipping each one costs you, most expensive first. That is deliberately not the order you perform them in, and the execution order is given after the list. Ordering them this way is the only honest way to decide what to do when a call comes in at four o'clock and you have time for two of the six.

1. Tell your contact you are escalating, before you escalate

This is the most expensive step on the list, and it is the only one whose loss cannot be recovered by doing it late. Every other failure here is repairable with a follow-up. A contact who learns from their own boss that their vendor went around them does not come back, and their replacement inherits the story.

The move is a sentence, delivered directly: "This needs a decision above your level and I would rather you take it up than me. If you would rather I write it to your safety officer directly, tell me and I will copy you on everything." That hands them the choice and it hands them the credit. Most of the time they will take it up themselves, which is the outcome you actually wanted.

What skipping it costs: the account, on a delay. Not the ticket, the account.

2. Put the finding in their system before you put it in anyone's inbox

Second most expensive, because unlike step 1 it is not repairable at all. You cannot backdate a work order. If the finding is not in the institution's own record with the date you found it, then six months later, when there is an incident and an investigation, your email folder is a vendor's uncorroborated claim and their ticket system is the file of record.

Write the finding into the work order in their computerized maintenance management system, their CMMS, in complete sentences and with the condition described rather than the remedy recommended. "Heat exchanger cell cracked, verified by inspection, unit removed from service" is a fact in their record. "Recommend replacement unit" is a quote, and quotes get filed as sales.

What skipping it costs: your evidence, permanently, in the one venue where it would matter most.

3. Move sideways to whoever owns the risk, not up to whoever owns your contact

Third, and this is the step that separates escalation from complaint. Every institution has someone whose job is the risk category your finding belongs to: an environmental health and safety officer, an infection control practitioner, a risk manager, a fire marshal, a compliance officer. That person has an independent budget path and an obligation that does not run through your contact's chain of command.

Going up the chain asks a busy administrator to overrule a subordinate on a technical question they cannot evaluate. Going sideways asks a specialist to evaluate a question that is exactly their specialty. The second one gets answered.

What skipping it costs: the escalation itself, plus your contact's standing, because the person above them will ask them why they did not handle it.

4. Restate the finding as one of their risk categories, not as a repair

Fourth. Institutions fund exposure, occupancy and compliance. They defer maintenance. The same cracked heat exchanger is a deferred capital item in one sentence and a combustion-products exposure in an occupied assembly space in another, and only the second one has a person whose job it is to act on it this week.

Do not exaggerate to get there. Restate accurately in their vocabulary: what the condition is, who is exposed, how many of them, and for how long per day. If the finding genuinely is not a safety or compliance matter, say so and let it go into the capital queue where it belongs. A shop that routes everything through the safety officer gets ignored by the safety officer.

What skipping it costs: the response. Your request stalls, correctly filed as a vendor asking for work.

5. Name the decision, the decision-maker, and the date it stops being cheap

Fifth, and expensive only in time, which is why it sits here rather than higher. Institutions do not have deadlines, they have calendars: a board that meets monthly, a fiscal year that closes, a procurement threshold above which a formal solicitation adds weeks. Your finding has to be attached to one of those.

State it as a date and a consequence, not as urgency. "A decision after the October board meeting means the earliest install is mid-December" is a fact anyone can check. "This is urgent" is a tone, and administrators are immune to tone.

What skipping it costs: a budget cycle. Recoverable, but only by waiting for the next one.

6. Close the loop downward, in writing, whatever the answer

Cheapest to skip today and the most compounding over a term. When the decision comes back, write it into the same work order: what was decided, by whom, on what date, and what the interim control is. If the answer was no, record the no and the interim control without editorializing.

That record is what makes your next escalation credible, and it protects the person who decided, which is the whole reason they will take your call the next time.

What skipping it costs: the next escalation.

The order you actually perform them in is 2, 4, 1, 5, 3, 6: write the record, translate it into their risk language, tell your contact you are escalating and offer them the first move, attach the calendar consequence, go sideways to the risk owner if your contact has not moved, then close the loop.

Worked example: a cracked heat exchanger in September

A district with 14 buildings. During a September preventive maintenance visit, a technician finds a cracked cell in the heat exchanger of the unit serving a middle school gymnasium.

Before anything else. A suspect combustion appliance in an occupied building does not get fired to demonstrate the problem. The unit comes out of service on the spot: the manual gas cock is closed, locked and tagged under 29 CFR 1910.147, which is the correct standard here because this is a stored and hazardous energy isolation rather than electrical utilization work, and any work inside the control cabinet is preceded by de-energizing and verifying dead under 29 CFR 1910.333(b)(2) with the live-dead-live sequence in NFPA 70E-2021, 120.5. A personal carbon monoxide monitor is worn by anyone in the space while a combustion appliance is firing. Who holds the authority to formally red-tag varies by jurisdiction and by whether the gas utility owns the meter set; what does not vary is that your own documented lockout, with the customer's written acknowledgment on the field ticket, is yours to do and to keep.

Step 2, the record. The finding goes into the district's CMMS work order the same afternoon: cell cracked, verified by inspection, unit isolated and locked out, gymnasium currently without heat. Condition, not recommendation.

Step 4, the risk category. Restated: a combustion appliance discharging into a space that holds up to 400 students during assemblies, used roughly 6 hours a school day. That is an occupancy exposure. It is now the safety officer's category as well as the maintenance supervisor's.

Step 1, the contact. A call to the maintenance supervisor, not an email: this needs a funding decision above you, I would rather you take it, and here is the calendar problem.

Step 5, the calendar. The board meets on the second Wednesday. Replacement lead time is about 4 weeks and install is about a week. Approved at the October meeting, the unit is running in early November. Missed, the next approval is the November meeting, which puts delivery in mid-December and running in the week before the holiday break, roughly six weeks of a gymnasium heated by nothing in the coldest stretch of the term. That is the sentence that moved it, and it moved it because it is checkable rather than dramatic.

Step 3, sideways. The supervisor took it up himself on day 3, so the safety officer was copied rather than petitioned. Three emails over nine days, two people added to the thread, approved at the October meeting.

Step 6, the loop. The decision, the date, the approver and the interim control, which was the gym staying out of use for assemblies until the unit was replaced, all written back into the same work order.

Contrast the shop on an adjacent district that took an identical finding straight to the superintendent on day one. They got a meeting that week and the unit approved. They also got a maintenance supervisor who stopped answering the phone; acknowledgment on their routine tickets went from same day to about a week, and they were not invited to bid the following year. They won the ticket and lost the term.

How to verify you got this right

Two checks, and the first one is uncomfortable.

Ask your contact, after the fact, whether the escalation made their job harder. Not whether the outcome was good. If the honest answer is that they got questioned by their own management about why it took a vendor to raise it, you skipped step 1 or performed it as a formality.

Then, three months later, open the work order and read it cold. If someone with no memory of the event can read that one record and see what was found, when, who decided what, on what date, and what protected the space in the interval, the escalation is complete. If any of those five are missing, the record will not hold up at a re-bid or an incident review, and neither will you.

References

• 29 CFR 1910.147, control of hazardous energy for mechanical isolation and stored energy sources
• 29 CFR 1910.333(b)(2), de-energizing and verifying electric utilization equipment; NFPA 70E-2021, 120.5, live-dead-live proving sequence
• See related: The Facilities Director and What They Are Actually Judged On, How an Institution Decides to Spend Money, What Changes When the Customer Has a Facilities Department