Enterprise-Grade Crisis Management for Travel Agencies
A typhoon disrupted 200 of Maya's agency's travelers across three corporate accounts on a single weekend in 2026. By the time the news hit Twitter, the agency had 90 minutes to look like a real operator before procurement officers at all three accounts started asking questions. The 90 minutes set the tone for the recovery week and the renewal conversations six months later.
The alert arrived at 7:14am on a Saturday. By 7:25am, Maya was in the agency's incident channel. By 7:40am, the operations center was lit up with three of the senior team already on calls. The 90-minute clock didn't start when the news hit Twitter. It started the moment Maya read the alert, and what happened in those first eleven minutes mattered as much as anything else that weekend.
This is the crisis playbook her leadership team built and stress-tested across that weekend. It's the operating manual for the moments at scale.
The First 90 Minutes Set Everything
The first 90 minutes of a crisis are when customers form their lasting opinion of how the agency handled it. Decisions made faster than that are usually wrong; decisions made slower than that are too late.
The agency's operations center activates within 15 minutes of an external event being confirmed. Three things happen in parallel. A named incident commander takes over decision-making (the COO or a designated deputy). The communications tree triggers, internal teams, suppliers, and corporate-account procurement contacts each get first-touch outreach within 30 minutes. The customer-impact roster is built, every traveler currently in flight, hotel, or activity in the affected region gets categorized by severity.
The impact roster isn't built from scratch during the crisis. It's a saved CRM query that filters every active traveler by region, trip status, and severity-of-impact. One click, and the roster appears. The query was built six months earlier during a tabletop exercise. The discipline of pre-building the roster query before you need it is what separates 8 minutes from 45 minutes on the same task.
By minute 90, the agency knows who's affected, who's been told what, and who's responsible for the next decision. That clarity is the difference between a crisis well-handled and a crisis that becomes a customer story.
The Roles You Assign Before You Need Them
The mistake most agencies make is figuring out roles during the crisis. The agency that looks calm in the first 90 minutes has rehearsed roles months earlier.
Maya's agency has six standing roles, each with a primary and a deputy.
Incident Commander owns all decisions during an active incident. Single point of authority. Doesn't directly handle individual travelers; handles systems and triage.
Communications Lead owns the message, internal, customer, and public. The tree has a defined first-touch SLA per audience.
Customer Impact Lead owns the affected-traveler roster and the per-traveler recovery actions. Reports to Incident Commander on aggregate status; manages a small team of customer-recovery specialists.
Supplier Liaison owns the relationships with affected suppliers, airlines, hotels, ground operators. They handle rebookings, refund negotiations, and supplier-side updates.
Finance Lead owns the cost tracking, both the immediate financial impact and the recovery budget. Authorizes any per-customer compensation above a defined threshold.
Service Continuity Lead owns the parts of the agency that aren't affected, making sure customer service for unaffected travelers doesn't degrade because the team is focused on the crisis.
The roles get rehearsed quarterly with a tabletop exercise. The first time you run through them shouldn't be the live event.
The Communication Tree
The communication tree has three branches and a defined SLA per branch.
Internal first. Within 15 minutes, the leadership Slack/Teams channel has a confirmed-incident message with what's known and what's being assessed. Within 30 minutes, the full agency staff has a brief that says what's happening and who customers should be referred to if they call.
Suppliers next. Within 30 minutes, the supplier liaison has direct outreach to the airline, hotels, and ground operators in the affected region. The relationships matter here, the supplier liaison knows who picks up the phone at 2am.
The supplier liaison's contact list isn't a spreadsheet. It's a Rolodex that's been built over years of small daily exchanges with each supplier. The 2am call lands on a phone that gets answered because the relationship was already there. Agencies that try to build the relationship during the crisis discover that the relationship doesn't exist yet, and the call goes to voicemail.
Customers and corporate accounts last. Within 60-90 minutes, every affected customer has a personalized first-touch from a named human (not an automated message). Corporate procurement contacts get a separate executive-level briefing within the same window.
The order matters. Customers calling the agency before the agency has reached out to them is the failure mode that destroys trust. Even if the message is "we're working on it and will know more in two hours," that proactive contact rebuilds the relationship.
What "Recovery" Actually Means
Recovery isn't refunds. Recovery is the customer leaving the experience telling other people that the agency handled the situation well.
The math is concrete. A traveler who experiences a disruption and feels well-handled has roughly the same lifetime value as a traveler whose trip went smoothly. A traveler who experiences a disruption and feels mishandled has roughly the same lifetime value as a customer who churns immediately.
The recovery actions that matter aren't the financial ones. They're the human ones, a named person who calls them by name, who knows the details of their specific situation, who acknowledges the inconvenience without minimizing or maximizing it. The financial gestures (rebooking at the agency's cost, compensation for missed activities, modest goodwill credit toward future bookings) matter, but they're support for the human gesture, not a substitute for it.
The agency that gets the human part right and the financial part adequate retains the customer. The agency that gets the financial part right and the human part missing usually loses them.
The Postmortem Rules
Within five business days of the crisis being resolved, the agency runs a structured postmortem. The rules:
Blameless. The point isn't who got it wrong. The point is what to change. Naming individuals is forbidden in the document; the action items name roles or systems.
Timeline-first. The postmortem starts with a precise timeline reconstruction, what happened, when, who knew what, who acted. Subjective interpretations come after the timeline is agreed.
Three buckets. Every observation goes into one of three buckets: things that worked well (preserve), things that almost worked (improve), things that didn't work (redesign). Action items come from the second and third buckets only.
Followups tracked. Each action item has a named owner, a deliverable, and a deadline. The next monthly leadership review checks progress on all open postmortem items. Items that age past 90 days without progress get escalated.
Maya's agency runs 4-6 postmortems a year. Most are minor incidents. The discipline of running them on small events is what makes the response to big events professional.
What Gets Pre-Built Before the Next One
The pre-built crisis assets the agency now maintains include:
Customer communication templates for the 8-10 most common incident types, weather event, supplier failure, geopolitical event, health incident, transport strike, hotel overbooking, lost-document situation, group separation. Each template is a starting point, not a final message, the comms lead personalizes during the event.
Supplier playbooks per major partner, who calls who at 2am, what each supplier's standard recovery offer is, what's negotiable, what's escalation-required.
Customer-impact roster query, a saved CRM report that filters all current travelers by region, traveler-type, and severity-of-impact. The query is one click during a live event.
Authorization thresholds, pre-approved per-traveler compensation amounts that the customer-recovery specialists can offer without escalating to Finance Lead. Above the threshold, Finance Lead approves; above a higher threshold, Incident Commander approves.
The pre-built assets compress the first 90 minutes of any future crisis. The work that took 30 minutes to gather information during the typhoon weekend now takes 8 minutes because the assets are pre-built.
The Renewal Conversation Six Months Later
The procurement officers at the three corporate accounts affected by the typhoon weekend brought it up in their renewal meetings six months later. All three brought it up positively. The phrase that came up in two of them was "you handled it like a real operator."
The renewals went through. Two of the three accounts increased volume the following year specifically because the crisis response had built trust beyond what a smooth year would have built. A handled crisis can compound a relationship faster than a year of routine service.
The opposite is also true. A mishandled crisis in year one shows up in non-renewal conversations 18 months later. The procurement officer rarely names the crisis as the reason; they usually cite "looking at alternatives." The crisis is the silent reason.
That's the long-term lesson. Crisis response isn't a cost center. It's one of the highest-leverage moments to either compound or lose customer relationships at enterprise scale.
Further Reading
- Multi-Country HR for Travel Agencies Operating in 3+ Countries. The HR architecture that makes cross-jurisdiction crisis response possible.
- Enterprise RFP Playbook for Travel Agencies. Crisis-response capability is now a scorable RFP criterion.
- Multi-Department Handoffs: Sales, Fulfilment, Service. The everyday-handoff discipline that crisis response leans on.
