Articles · Travel
More enquiries but no more sales? The tours fall through after payment
12 October 2026
The advertising is working and enquiries are well up, but no more tours are being sold. The owner of a travel agency explains it the way anyone would: the managers are buried in tour shortlists, reply late and never get the client as far as paying. A vendor has a product ready — an AI assistant that puts the shortlists together itself. The owner is already wondering whether three managers would then do instead of four.
One way to check is to follow a single enquiry along its whole route, from the shortlist to the tour operator's answer. We will take the route at Yuzhny Meridian travel agency. The agency is invented: companies like it are what we test our audit agent on, and the agent wrote the report on its data. In thirteen weeks it took 1,578 enquiries and sold 169 tours.
The shortlist: twenty minutes, not half a day
The owner believed shortlists took the managers half a day. The senior manager's figure is different: 15–20 minutes a shortlist, six or seven a day. Most time goes on looking after tours already sold — documents, additional payments, flight changes: three to four hours a day for each manager.
Shortlists take about 17% of the managers' time. The assistant would take up to 110.8 hours a month off the four of them, but that is the time of salaried staff, not money: not one payment the firm makes would change. Nor would three managers be enough: the after-sales work, calls, bookings and duty shifts all remain.
Whether the managers handle clients well after the shortlist goes out, the report does not say. There is no log of calls and messages, so "they sell badly" can be neither confirmed nor ruled out. Further along the route, though, is a place where sales are lost for certain.
Payment: the client pays, and the enquiry stops moving
The client has chosen a tour and paid. For the manager, the deal is done. But at this agency a tour can be booked with the operator only once the enquiry is marked "Payment confirmed", and only the accountant makes that mark. She works two days a week, Monday and Thursday, from 10:00 to 15:00. The online-banking software is on the office computer.
Take a payment that arrives at four on a Thursday afternoon. The accountant left an hour ago. On Friday, Saturday and Sunday the money sits in the agency's account and the manager may not book: there is no mark. It will appear on Monday.
The records show this without anyone being asked: all 189 first bookings on paid enquiries were made on a Monday or a Thursday. The typical (median) gap between payment and booking is 1.9 days, and it can be as long as four. The owner was sure payments were confirmed within the hour.
Booking: the operator answers in an hour and a half, but a price lasts a day or two
The operator is not the slow one: it responds within an hour and a half. But it is responding to a booking that reached it late. While the payment waits, the places sell out or the price goes up; as the senior manager says, an operator's price "lasts a day or two".
The numbers agree. When the booking is made within a day of payment, the operator fails to confirm 4 bookings out of 59. With a wait of two to three days it is 16 out of 54. Beyond three days, 12 out of 25.
Of the bookings made within a day of payment, the operator failed to confirm 4 out of 59. Beyond three days, 12 out of 25.
The end of the route: the client pays extra or takes the money back
In all, the operator did not confirm 43 of the 189 first bookings. On 24 enquiries the tour was rebooked, usually at a higher price, with the client paying the difference. On 19 the client took their money back. The commission on those 19 bookings comes to 3,586,700 KZT over thirteen weeks.
These were tours already sold. The advertising and the manager's time had been spent on these clients; they had chosen a tour and paid. It all came apart at a step the owner took for an hour's work.
From the inside a loss like this cannot be seen, because everyone involved is in the right. The manager made the sale. The accountant made the mark on her working day. The operator answered in an hour and a half. The wait between payment and booking belongs to nobody, and until it is counted from the dates, a refund looks like a client who changed their mind.
"More enquiries, the same sales" is explained by the advertising
Instagram advertising brought in 662 enquiries, and 21 turned into sales. That is 3.2%, against 16.2% from all other sources. According to the manager, people there ask "so how much is it?" and vanish. The new stream buys five times less often than the old, so enquiries grow faster than sales with the same managers.
Even so, the advertising pays for itself: at a spend of 280,000 KZT a month, a deal costs about 40,000 KZT and leaves 100,843 KZT of commission after the manager's share. There is no reason to switch it off. The report advises showing a "from" price in the ads and asking about dates and budget in the very first message. As for "twice as many enquiries", there is nothing to check it against: the agency has no data from before the advertising.
The first step costs 3,000 KZT a month, and it is not the assistant
- The shortlist assistant — 250,000 KZT plus 60,000 KZT a month. It speeds up a step where the enquiry is not held up, and leaves alone the point where the money is lost.
- Automatic reminders to clients — 40,000 KZT plus 10,000 KZT a month. Not now: without an event log nobody can see whether the managers get back to clients themselves. The log has to be requested from the supplier of the enquiry system.
- The first step — confirm each payment on the day it arrives. The senior manager gets read access to the bank statement and the right to mark payments as confirmed; the bank charges 3,000 KZT a month. And someone is named to do it when she is away.
- If the daily check does not stick — link the bank to the enquiry system: a one-off 90,000 KZT plus 15,000 KZT a month. Not both at once: they remove the same wait.
An agent has a place at this agency, but later. Once there is an event log, an agent can find enquiries with no next step and prompt both the manager and the client; at present it has nothing to work from. For now a free fix will do. There are 894 enquiries sitting in "On hold": clients who go quiet are moved there to avoid closing the enquiry as lost and having to give a reason. So a reason becomes mandatory for that move. Refunds, conversations about paying extra and booking with the operator stay with people in any case.
The report promises no savings: none of the changes removes a payment. The aim is to stop losing tours already paid for; how many can be saved, a comparison after four weeks will show.
Three numbers a travel agency can pull today
- The time from payment to booking. Take the payment date from the bank or payment-service statement and the date of the first booking from the enquiry system, for every paid enquiry in the quarter. Look at the median and the longest case. Anything over a day is a warning sign. So are bookings bunched into two or three days of the week, like Monday and Thursday in the example: then ask who marks payments as confirmed, and at what hours that person is in.
- The share of bookings the operator did not confirm — prompt ones and late ones separately. The same enquiries, the status of the first booking. In the example, 4 out of 59 against 12 out of 25. If late bookings are refused several times as often, you are looking at the price of the wait. Alongside, note how many paid enquiries ended in a refund.
- Sales per hundred enquiries, by source. The report by source in the enquiry system. In the example, 3.2% for one channel and 16.2% for the rest. If a new channel brings many enquiries and few sales, that explains "more enquiries but no more sales" without bringing the managers into it.
If your first number is a matter of hours and the operator's refusals do not depend on the wait, your case is not the one in the example. Then the question about the managers is worth going back to — with the log of calls and messages in hand, not an impression.
Want the same look at your own company? The audit is free for now — it is an experiment. Show us how your work runs and get a report: where time and money are lost, and where AI would help and where it would not.
Request an auditSee sample reports
“Audit” here means a review of how the company works, not an audit of financial statements. The companies in the examples are invented; the reports on them were written by our AI agent.