Don’t Show Me Your Emergency Manual

Show Me What Happens at 3 A.M.

Issue 12 | October 2026

A periodic publication by 3AM Hospitality Consulting.

It is 3 a.m. and your hotel is full when the shaking starts. The power goes out, guests begin pouring into the corridors and somewhere in the building a fire alarm activates. Mobile phones are unreliable and nobody knows yet whether the building is structurally safe.

Your General Manager isn't there. So who is actually running the hotel?

I've spent more than 30 years in hotels and, unfortunately, I've had enough opportunities to find out. Earthquakes, fires, floods, prolonged blackouts, food-safety incidents, civil disturbances and more evacuations than I care to remember. None happened exactly as the manual said it would.

I've learned that trying to write an SOP for every possible crisis isn't the answer. What matters is preparing the organization that will have to deal with it.

The big one isn't just an earthquake

PHIVOLCS has long used a magnitude 7.2 West Valley Fault earthquake as a planning scenario for Greater Metro Manila, with modelling showing potentially severe casualties and widespread damage. These are scenario estimates, not predictions, but they give us some idea of the scale of disruption we should be preparing for. [1]

A hotel also has a very different problem from an office that can simply close and send everyone home. At 3 a.m. we may have hundreds, perhaps thousands, of people asleep inside our buildings. And the earthquake itself may only be the beginning. Fire, casualties, power and water failures, blocked roads, overwhelmed hospitals and communications failures can all follow.

That's what worries me about the Big One. Not simply the earthquake, but what happens next.

I've been in the room when the building started moving

Years ago, I was in my hotel room with my wife when a major earthquake struck. The power went out, guests were panicking in the corridors and I could hear the building itself creaking. My training told me what to do. My wife had other ideas about being left under a table while I disappeared downstairs, so we evacuated together.

Once I reached the lobby, I shifted from husband to hotelier. We declared the emergency, established our command centre and began evacuating the hotel—all 640 rooms. We had staff lists and an in-house guest list, but that wasn't enough. Restaurants and bars had to be cleared and rooms physically swept. Some guests didn't want to evacuate. My instruction was simple: this is an emergency; evacuation isn't optional.

Then the generator kicked in, which created a problem I hadn't expected. Guests saw the lights come back on and started asking to return to their rooms. But lights don't mean a building is structurally safe.

We waited roughly four hours for structural clearance. We were lucky. Our Chief Engineer knew a structural engineer who could get to us. We had to wake him. It worked, but luck isn't a contingency plan.

During the Big One, that engineer may be looking after his own family. Roads may be blocked and every hotel, condominium and office building around you may be making exactly the same call. So who is your second call? And your third?

There was another small incident that night that taught me almost as much. I discovered we had a traffic jam leaving the hotel because Security and the cashier were still collecting parking fees. They were doing exactly what they had been trained to do, and that was the problem. I radioed down: stop collecting, open both gates and let everyone out.

In a genuine emergency my rule has always been simple: protect people, make the decision and explain it later. The authority to act needs to be agreed beforehand, and people need to practise using it. Nobody should be discovering the limits of their role while guests are trying to get out.

Karl Weick made an important observation about organizations operating under extreme pressure. When familiar structures begin to break down, procedures alone aren't enough. People need to communicate, trust each other, improvise and make sense of what is happening together. [2]

An emergency manual tells people what to do when the situation resembles the manual. Resilience is what happens when it doesn't.

The hotel resilience cycle

Over the years I've come to think about crisis preparedness in four stages: Survive, Stabilize, Sustain and Recover.

Survive

The first hour is about life safety, accountability, command and immediate threats. If the General Manager isn't there, the Duty Manager needs the training and delegated authority to take command immediately. The team should already know who leads and who takes over if that person is unavailable.

That also means the DM shouldn't be running off to fight the fire. Someone else leads the emergency response team while the person in command maintains the overall picture. Operations, Engineering, Security, Emergency Response, HR, Guest Care and Communications each need somebody responsible—and somebody who can replace them.

If your emergency plan depends on the GM being there, you don't have an emergency plan.

Stabilize

Once the immediate danger is under control, the question changes from whether people are safe to whether we have control. This is also when the information crisis begins. Guests are posting, families are calling, employees are messaging, authorities need information, owners want answers and journalists may already be outside.

Communications belong in the command structure from the beginning.

Years ago, I went through some fairly brutal media training, with ambush interviews, cameras, lights and an actual television personality deliberately putting us under pressure. One simple framework stayed with me: Acknowledge. Act. Add Perspective.

Acknowledge what happened and the people affected. Explain what you are actually doing, then add whatever verified context helps people understand the situation. Perspective is not spin. If you don't know something, say so. [3]

I had to apply the same principles during a food-safety incident when close to 50 doctors attending an event became ill. Our nurse spotted the pattern and, because we retained food samples, we could immediately start testing. The normal microbiological tests came back clean, but after discussing the symptoms with the laboratory manager we ran additional tests that identified elevated histamine associated with fish supplied for the event.

We shared everything with the organizers and focused on looking after those affected while we worked through what had happened. Our team did an outstanding job, and the company remained one of our major clients.

The clean standard results would have been a convenient place to stop. They hadn't explained why people were ill. Continuing the investigation and sharing what we found mattered more than having reassuring results to announce.

Sustain

The earthquake has stopped, the immediate fires are under control and people have been accounted for. Now comes the question I suspect many emergency plans haven't properly answered: can we look after ourselves?

For a large hotel, I would stress-test that question over 14 days. This isn't a government requirement and it doesn't mean putting fourteen days of everything into a storeroom. It means understanding what happens if normal city infrastructure and supply chains cannot be relied upon, including when we would need to relocate guests or close. Fourteen days is a planning horizon, not a reason to remain in an unsafe building.

I learned why this matters in Kuala Lumpur. We were told municipal water would be cut to our area for 24 hours and initially thought we were prepared. Then Engineering did the actual consumption calculations. Because of the cooling towers, our water wasn't going to last anything close to what we had assumed.

We scrambled to arrange tanker deliveries and were fortunate that the interruption was relatively localized. The tankers kept coming. Had the whole city needed those same tankers, we would have had a very different problem.

The useful question is how long the water actually lasts with the hotel operating. [4]

Recover

Eventually the question becomes how to get back safely, and sometimes the correct answer is not to reopen quickly.

At another hotel, replacing obsolete transformers meant closing for roughly two weeks. That interruption was necessary to do the work properly. Business continuity doesn't always mean staying open. Sometimes the best decision for the business is to close, fix the problem properly and come back safely. [5]

For owners, think of it as insurance

Preparedness costs money, but I think owners sometimes ask the wrong question. What is the ROI on another radio, another drill, a structural-engineering retainer, more water storage or additional redundancy?

I wouldn't look at it that way. I would think about resilience much more like insurance. You hope you never need it and most of the time it sits quietly in the background. Compared with the alternative, however, it can be remarkably cheap.

The first concern is people. The commercial consequences can also last long after the physical repairs are finished. Corporate accounts can leave, employees remember how leadership behaved, and owners, insurers and regulators start asking difficult questions.

From an investor's perspective, this is asymmetric risk. The sensible cost of preparedness is relatively small; the downside of getting it badly wrong can be permanent.

Stop practising perfect emergencies

The problem with many hotel drills is simple: people don't take them seriously enough. Everyone knows it's a drill and everyone knows what's supposed to happen. Too often we practise fire this year and then practise fire again next year.

I'm a big believer in tabletop exercises for hotel leadership. Alongside drills that practise the response, they give the team time to work through decisions. Put them around a table and then make things progressively worse.

I would expose different shifts to different scenarios throughout the year and conduct at least one serious leadership tabletop annually. Then, after every meaningful exercise or real incident, conduct an after-action review: what was supposed to happen, what actually happened, why, and what are we going to change? [6] Record what needs to change, give someone responsibility for fixing it, and test it again.

When the hotel becomes something else

There is one other thing I remember from that earthquake. Hundreds of guests were outside for hours waiting for the building to be cleared. Once our colleagues were accounted for and the essential emergency positions covered, the rest of the team went back to doing what hoteliers do: looking after people.

We gave out water and blankets, checked on guests and kept them informed. The building temporarily couldn't perform its normal function, but the people still could.

After an evacuation, hospitality doesn't stop. It changes form.

Luxury at that moment isn't thread count or lobby design. It is water, a blanket, a working toilet and reliable information. It is knowing somebody is in control, noticing that an elderly guest needs help and having someone there who can say, “We're looking after you.”

Perhaps that is the simplest definition of duty of care.