Following One Trip Through Five Tools: A London–Nice Charter and the Seven Places It Gets Retyped
Seven times, a person retypes the same London–Nice charter before its invoice is settled, and in the model below that retyping absorbs about 45 hours of broker time in a 40-trip month. The figures are an illustrative model built on stated assumptions, not villiersOS customer results.
Take one trip: Farnborough (EGLF) to Nice (LFMN) on a Friday morning, back on Sunday evening, four passengers with luggage, on an Embraer Phenom 300E. The aircraft seats up to eight and covers the sector in about 1 hour 45 minutes of block time. Light jets of this class typically price at £7,500 to £10,500 per sector, so the return, built as two one-way sectors, runs £15,000 to £21,000.
The broker works across five tools: quoting, scheduling, compliance, billing and a CRM. Facts pass between them by hand at seven points:
- Enquiry to quote. Route, dates and passenger count move from the client's email into the quoting tool.
- Operator quote to client quote. Tail number, sector price, fuel and handling assumptions are retyped, with the broker's margin added.
- Accepted quote to schedule. Tail, times, FBO and passenger count are entered again in the scheduling tool.
- Manifest to compliance. Names, passport numbers, dates of birth and nationalities travel from the client's reply into the compliance tool.
- Schedule to duty check. Departure times, arrival times and crew names are typed a third time so duty limits can be tested.
- Handler and fuel actuals to billing. The final fuel uplift, handling and landing charges are keyed against the quoted price to build the invoice.
- Invoice to CRM. Final price, aircraft flown and the passengers' preferences are logged for the next enquiry.
The tail number is typed at four of the seven points. None of the five tools can see the others, so none can catch a transposed digit.
Points 3 and 6 deserve particular attention. Point 3 fixes the aircraft and times that every later tool inherits, and point 6 is where the quoted price meets the operator's actual charges, so an error at either one reaches the client's invoice.
The Arithmetic of Re-Keying: Minutes per Handoff, Error Rates and What a 40-Trip Month Costs in Broker Time
Sixty-two minutes a trip is the starting assumption, and it is deliberately modest. The assumed minutes for points one to seven are 6, 10, 8, 12, 6, 15 and 5, which sum to 62. Point six runs longest because reconciling handler invoices to the quote means reading line items, not copying a field.
Across 40 trips that is 2,480 minutes, or 41.3 hours. Add late changes: assume 30% of trips change after acceptance, which is 12 trips, and each change is retyped in four places at 5 minutes apiece. That is 20 minutes a trip and 240 minutes, or 4 hours, a month.
The total is 45.3 hours, more than a quarter of one full-time coordinator at 160 hours a month. At an assumed loaded cost of £45 an hour (the author's estimate for salary plus overhead), that is about £2,040 a month.
Errors cost more per incident than minutes do. Assume 3% of handoffs introduce at least one wrong field; seven handoffs across 40 trips is 280 handoffs, so about 8.4 errors a month. The usual types are a transposed tail digit, a passport number off by one character, a local time keyed as UTC, a handling fee left out of the quote, and a quote total that no longer matches the operator's invoice.
Assume three in four are caught inside the workflow, at a cost already counted above. The remaining 25%, about 2.1 a month, escape to the client, operator or handler, at an assumed average of £250 each for credit notes, surcharges and corrected paperwork. That adds about £525.
The monthly total is £2,565, or about £64 a trip. Set against an assumed 10% broker margin on the £18,000 midpoint return, which is £1,800 a trip and £72,000 across 40 trips, re-keying consumes about 3.6% of gross margin, or £30,780 a year.
The average understates the exposure. One escaped error can erase a large share of a single trip's £1,800, as the next section shows.

The Late-Change Failure: When a Tail Swap Updates the Schedule but Not the Quote, the Crew Duty Check or the Invoice
At 15:40 on the Thursday before departure, the operator swaps the Phenom 300E for a Citation XLS+, which seats up to nine, and retimes the Friday departure from 09:00 to 07:30 local. The scheduling tool takes the update because the operator's notification lands there. Nothing else hears about it.
The quote still shows the Phenom 300E at £8,200 for the outbound sector, while the operator now bills £9,400. The £1,200 difference is 67% of the £1,800 margin on the trip (1,200 divided by 1,800).
The compliance record carries the old registration, so the general declaration sent to the handler at Nice names the wrong aircraft. The duty check was run against a 09:00 departure and the original crew pairing. The new aircraft brings a different crew and a start 90 minutes earlier, and neither was tested.
The invoice is then generated from the quote. The client is billed £8,200, the broker owes £9,400, and the gap surfaces only when the operator's invoice arrives days after the trip. The paperwork gap surfaces sooner, and at the worse moment: on the ramp.
Training does not remove this failure. Each tool is correct about what it was last told, and the trip as a whole is wrong.

Why Integrations and Zapier Glue Only Move the Gap: Field Mapping, Sync Lag and No Owner for the Mismatch
Charter operations software integration done with point-to-point connectors tends to fail in three predictable ways. The first is field mapping. The quoting tool stores one price field, while billing splits sector price, handling and fuel into three; a zap maps the total into one and leaves two blank, so the invoice builds from partial data.
Five tools have ten possible pairs, and each pair needs its own mapping, maintained by whoever built it. A renamed field in any one tool breaks a connection that nobody is watching.
The second is sync lag. Assume a 15-minute polling interval, an illustrative figure. The 15:40 tail swap reaches the compliance tool at 15:55, and a duty check run at 15:45 passes on stale data.
The third is ownership. When a connector maps wrongly or fails quietly, the error sits between two tools, and no tool owns the mismatch. The broker learns of it from a client or a handler.
Even a well-built zap carries no judgement. If handling at LFMN is quoted at £900 and invoiced at £1,350, the connector copies both numbers faithfully and raises no flag.
Joining up quoting, scheduling and compliance software this way makes the seven retyping points faster. It leaves unchanged the number of places where one fact exists as several copies, and that number is what produces the errors.
One Trip Record, Five Agents: How a Shared State Closes the Gaps, and the Handoffs a Human Should Still Own
Tail number, price, times and manifest each live in exactly one place in this design. In villiersOS, the quoting, scheduling, compliance, billing and CRM agents work from a single trip record. The quoting agent writes price and aircraft; the scheduling agent reads the aircraft and writes the times; the compliance agent reads the manifest, registration and times; the billing agent reads quoted and actual costs; the CRM agent reads the outcome.
Nothing is retyped because nothing is copied. The seven points collapse to one capture, the client's email, and one set of actuals from handlers.
Replay the Thursday swap. The scheduling agent writes the new aircraft to the record at 15:40. The quoting agent sees that the sector price no longer matches and flags a £1,200 variance. The compliance agent re-reads the registration and regenerates the declaration, the duty check re-runs against 07:30, and the invoice draws on the corrected quote. One write has four consequences, and they arrive in minutes rather than after a polling interval.
Under the model's assumptions, this removes the 41.3 hours of base re-keying, the 4 hours of late-change work and most of the 8.4 monthly errors, since a field with one copy cannot disagree with itself. It does not remove the work of deciding.
A human should still own four handoffs:
- negotiating price and aircraft with the operator;
- approving a tail swap that raises the client's price before the client is told;
- resolving a compliance flag on a named passenger;
- calling the client when the trip changes materially.
A simple threshold keeps this workable. Any variance above 5% of the sector price, which is £410 on £8,200, goes to a person before the client's quote changes; that level lets a small fuel difference pass but never a change of aircraft, since the £1,200 swap is about 15% of the sector price.
To test this on your own book, time the seven points on ten real trips, multiply by monthly volume at your loaded hourly cost, and add the escaped errors. If the total sits near the 3.6% of margin modelled here, specify the shared record first around the tail swap, because that is where one missed update costs most.





