The 90-Second Window: What Happens the Moment a Request Hits villiersOS
A request submitted through villiersOS at 14:02 on a weekday afternoon can return a ranked shortlist of three aircraft, complete with tail numbers and firm hourly pricing, by 14:03:30. That is not a marketing round number; it is the median turnaround logged across live requests in the platform, against an industry-typical phone-and-email quoting cycle of twenty to thirty minutes per operator contacted.
The compression happens in a fixed sequence. First, intake parses the raw request (departure airport, destination, date, passenger count, any stated aircraft preference) into a structured query. A request for "four people, London to Nice, next Friday" becomes a normalised object: EGLL or a London-area alternate, LFMN, a passenger count of four, a departure window, and a category filter if one was specified.
Second, that structured query becomes a request-for-quote (RFQ) object, timestamped and logged, before it ever reaches an operator. This is the part clients never see but that matters most for accountability: every RFQ has an audit trail, so a broker can later show exactly which operators were asked, when, and what they returned.
Third, the RFQ fans out simultaneously rather than sequentially. This is the structural reason villiersOS charter software compresses a half-hour phone round into under two minutes: no operator is queried, then waited on, then the next one dialled if the first is slow. All qualifying operators receive the request at the same instant, and responses stream back asynchronously as each one clears its own availability check.
Fourth, as responses arrive, the scoring layer begins ranking them in real time rather than waiting for every operator to reply. By the time a broker opens the request, the shortlist is already ordered, and late responses simply insert themselves into the ranking rather than triggering a re-run.
The Operator Fan-Out: How the Network Gets Queried and Ranked
Not every operator in the network gets every RFQ. The fan-out is filtered before it is sent, not after: an operator whose fleet has no aircraft with the range to cover a London-to-Dubai sector, or whose certification does not cover the departure country, is excluded at the query stage rather than surfaced and then rejected.
The villiersOS charter software maintains live connections into an operator base of several hundred approved carriers, cross-referenced against category (light jet, midsize, super-midsize, heavy, ultra-long-range), home base proximity to the requested airports, and current certification status, including EASA and FAA air operator certificates where relevant. A request for a four-passenger EGLL–LFMN sector on a midsize jet, for instance, only goes to operators holding aircraft such as a Cessna Citation Longitude or comparable midsize type based within a repositioning radius that keeps the quote commercially sensible; an operator whose nearest aircraft sits in Malta gets excluded, because the empty positioning leg would make the quote uncompetitive before it is even generated.
Responses come back with availability, aircraft type, tail number where the operator discloses it, and price. At this stage the list is a straightforward availability match; nothing has yet been ranked on anything other than whether the aircraft can physically do the trip. That ranking is the next stage, and it is where the platform starts to differ meaningfully from a simple broadcast-and-collect quoting tool.
A London–Nice sector on a midsize jet such as the Citation Longitude typically returns quotes in the region of £18,000 to £24,000 one way at current market rates, depending on positioning distance and date. Seeing that range against three or four operators within the same ninety-second window, rather than piecing it together from separate phone calls across an afternoon, is the practical value the fan-out delivers to a broker working a live client request.
Scoring Beyond Price: Reliability, Cancellation History and Aircraft Age in the Algorithm
Price alone is a poor ranking signal, because the cheapest quote on a given sector is disproportionately likely to come from an operator with a below-average on-time record or a fleet aircraft older than the client would expect for the price bracket. The scoring model weighs several factors against each other rather than sorting by price ascending.
On-time performance and cancellation history carry real weight. An operator with a 96% on-time departure rate and a cancellation rate under 1% over its trailing twelve months scores materially higher than one at 88% and 4%, even at an identical headline price, because a late-cancelled charter is a reputational cost to the broker as well as a logistics problem for the client. Safety accreditation, tracked against third-party audit bodies such as ARGUS and Wyvern, feeds into the same weighting.
Aircraft age and cabin condition matter too, particularly at the upper end of the fleet. A Gulfstream G650ER delivered in the last three years scores differently from a similarly specified airframe approaching its fifteenth year, even though both meet the range and passenger brief; the algorithm treats airframe age as a proxy for cabin refurbishment cycle and dispatch reliability, not as a hard exclusion. Operator responsiveness, measured as time-to-quote on prior RFQs, is folded in as a smaller but non-trivial factor, because an operator that takes six hours to confirm availability is a poor fit for a platform built around a ninety-second window.
The result is a composite score, not a single number a client ever sees directly. What reaches the next stage is an ordered shortlist where the top result balances price against reliability, rather than simply being the lowest bid.
The Human Checkpoint: Why Every Quote Still Passes Through a Broker Before It Reaches a Client
No ranked shortlist from villiersOS charter software goes to a client without a broker opening it first. This is a deliberate architectural choice, not a placeholder waiting to be automated away. The algorithm has no visibility into a client's stated preferences from a previous trip, a standing instruction about a specific operator they have flown badly with before, or a diary conflict that makes a particular departure slot impractical even though the aircraft is available.
A broker reviewing the shortlist can reorder it, add context the system has no access to ("this operator's cabin crew were excellent on the client's last Nice trip"), or pull a quote entirely if something in the operator's response looks off, an unusually fast turnaround on a normally slower carrier, for instance, which experienced brokers read as a signal worth a phone call before it goes to a client. The system surfaces the ranking; the broker decides what actually gets sent.
This checkpoint also does the quiet work of margin and positioning that no scoring algorithm should be making unsupervised: adjusting a quote where a repositioning cost has been miscalculated, or flagging to a client that a marginally more expensive option on the list is worth the difference for reasons the client would only value if someone explained them. Villiers has been explicit internally that the fan-out and scoring stages exist to compress the mechanical part of sourcing a quote, not to replace the judgement that turns a shortlist into a recommendation a client trusts.
Empty Legs as Live Inventory: How Repositioning Flights Get Surfaced in Real Time
An aircraft flying back empty after dropping a client is treated by the platform as active inventory the moment the operator logs the return sector, not as a special category checked separately. A Bombardier Global 7500 repositioning from LFPB Paris Le Bourget back to EGLL after a one-way charter becomes a live entry the instant the operator's system confirms the empty return, and it is scored against any open request whose date and route window overlap it.
Because the matching runs continuously against incoming RFQs, an empty leg can appear in a shortlist within the same ninety-second window as any other quote, priced well below a purpose-flown charter on the same city pair because the operator only needs to cover the marginal cost of the return sector rather than the full round trip. A client requesting Paris to London on short notice might see a standard midsize quote alongside an empty-leg option on a larger aircraft at a comparable or lower price, purely because a repositioning flight happened to align with their dates.
The trade-off is fixed timing: an empty leg is tied to the operator's original booking, so the departure slot is not negotiable in the way a dedicated charter is. That constraint is why empty-leg inventory is surfaced as one ranked option among several rather than promoted as the default; a broker still weighs whether the fixed slot suits the client before recommending it over a flexible, purpose-flown quote.
The same fan-out, scoring and repositioning-inventory architecture is not aviation-specific in principle. The underlying pattern, a fragmented supplier network queried simultaneously, ranked on more than price, and checked by a human before a client sees it, applies equally to any high-value, low-frequency booking category with comparable supply fragmentation, from yacht charter to high-value vehicle rental, though villiersOS itself operates only within private aviation today. What makes it work for Villiers specifically is the same thing that would make it work anywhere else: the algorithm handles volume and speed, and a person still signs off on judgement.
