The booking a travel agency lost last month probably wasn't lost on the itinerary. It was lost in the gap between the traveler's message and the reply — and not for the reason most people assume.
Everyone frames fast follow-up as a race for first place: reply first, get in line ahead of the other three agencies. My read is that the race framing undersells it. A stranger sending a first inquiry has no way to test whether you know the destination, so the first thing they can grade is the reply itself. The first reply is not the start of the service — it is the only sample of the service they have when they choose.
Shrinking booking windows and waiting strangers
This should sting more for an agency that competes on service than for one that competes on price. Your whole pitch is that a person who knows the place will handle it properly, and the first evidence a stranger gets of that claim is a timestamp. It also matters more than it used to, because the deciding window keeps shrinking: Criteo data reported by Skift found that over 30% of U.S. travelers now finalize plans within two weeks of departure, and in a Travel Weekly reader survey conducted in March, the trend of close-in bookings appears to be continuing. When a trip gets booked two weeks out, a two-day reply eats a real share of the time they spend deciding.
The sharpest measurement of this reflex is not in travel at all — it is in the wider business of handling online inquiries, which has studied the clock far more closely than our industry has. In its March 2011 issue, Harvard Business Review published research on how companies handle potential customers' online queries, and its verdict was that most companies are not responding nearly fast enough. That verdict is fifteen years old, which is the interesting part: nothing that has happened to buyers since 2011 has made them more patient, and a leisure inquiry inherits the problem in a harder form, because there is no purchase order behind it to stop the sender from messaging two other agencies while yours sits unread. My own read is that the number to watch is not how long people say they are willing to wait, but how long a stranger waits before they stop thinking of themselves as your client — and that clock starts running long before the itinerary is even a question.
Automate the acknowledgment, not the advice
The wrong conclusion is that if speed wins, you should automate the answer. Asia's large travel platforms have argued the adjacent question in public, and the mechanism is the same one sitting in your inbox — a stranger waiting. At the Skift Asia Forum, executives from Traveloka and Klook argued travelers are most interested in high-quality service at the right time, even if it comes from AI, and when the moderator asked whether "high touch" requires a human concierge, almost no one in the crowd raised their hand. What that panel describes is the answer, though — not the recommendation.
That distinction is where I'd draw the line in your shop. As Daniel Dines puts it in The Work That Remains: "Some work can be mostly right: a draft, a brainstorm, a summary a person will check. Other work cannot." Acknowledging an inquiry in ninety seconds and asking for destination, dates and group size can be mostly right. Deciding where a family of six should actually go in August is the part with your name on it. Speed on the wrapper, judgment on the contents.
The first reply is not the start of the service — it is the only sample of the service they have when they choose.
Log seven days of timestamps before you change anything
For one week, write down two numbers for every inquiry: when it arrived and when a human first replied. Nothing else — no scoring, no CRM project. At the end of the week, sort the list by hour of arrival. The evenings, weekends and lunch hours are where the gap will be, and you cannot argue with your own column of numbers.
We've watched this play out with our own clients. A growing travel agency: 5× monthly sales — 8 to 40+ in 90 days. The case is here if you want to see how it was built: https://brandrstudio.com/our-work/travel-agency-ai-sales See the case →
Get this every week. One issue, one idea, sources you can check. No pitch.
Sources
- Skift — What Winter Travel Booking Trends Reveal About 2026 Demand
- Travel Weekly — Travel trends in 2026: Mixed volume and shorter booking windows
- Harvard Business Review — The Short Life of Online Sales Leads (March 2011)
- Skift — Travelers Don't Care If It's AI or Human — They Just Want the Answer Now
- [The Work That Remains: Human Judgment, AI, and the Architecture of the Next Enterprise — Daniel Dines](library:The Work That Remains: Human Judgment, AI, and the Architecture of the Next Enterprise)