Farnborough
Nice
Why Timer-Based Chasers Fail on a Live Charter Enquiry
A client who has viewed four quotes on the dashboard and then spent forty minutes on the phone with your desk does not need a reminder email at nine the next morning. A timer sends it anyway, because a timer cannot see the phone call.
The standard broker chaser is a schedule: send the quotes, wait 24 hours, chase; wait 72 hours, chase again. It is cheap to build and easy to explain. It also treats silence as one condition, when on a live charter enquiry silence has at least four different causes: the client is comparing quotes, is waiting on a colleague or PA, has booked elsewhere, or has not opened the email at all.
Each cause needs a different response. A client who is comparing needs space. One who has not opened anything needs a nudge by another route. A client who has been quiet for ten days may need a person to ask, plainly, whether the trip still exists. The timer sends the same email to all of them.
The cost is easy to size, using an illustrative desk and our own estimate rather than a published figure. Suppose a desk quotes 40 enquiries a week and a timer chases each one twice: that is 80 automated emails a week. If one in ten lands on a client who is already in conversation with the desk, that is 8 badly timed emails a week, or 416 a year (8 × 52). Each one goes to someone considering a sector that typically costs £9,000 to £14,000 one way on a light or midsize jet between London and Nice.
That is why a private jet charter quote follow-up needs a decision before it needs a draft. The rest of this piece covers the decision layer inside villiersOS that Villiers runs on its own charter desk.
The Active-Thread Check: What the Agent Reads Before It Writes a Word
The first thing the agent does with an unanswered quote is look for evidence that the conversation has moved somewhere it cannot see from the quote record alone. It reads the trip's email threads, including messages from the client, from a PA or assistant on the same trip, and from operators. It reads the call history against the same trip. Only then does it consider drafting.
The reasoning is simple. A broker working a charter enquiry spends much of the day in two channels, email and telephone, and the quote record knows about neither. If the client rang yesterday afternoon to ask about catering on a Phenom 300E, the enquiry is active, whatever the quote status says.
The second input is engagement. The agent reads the dashboard view count, meaning how often the client has opened their quote page, and the quote count, meaning how many options are actually on the table. These two figures separate cases that look identical on a timer. A client with five quotes and six dashboard views is working through a comparison. A client with one quote and no views may never have seen the email.
Here is how that plays out on a single trip. Four quotes went out on Monday for Farnborough (EGLF) to Nice (LFMN). By Tuesday morning the client has opened the dashboard three times and replied to a colleague's forwarded message on the same thread. A timer would chase. The agent sees a live thread and reads high engagement, so it leaves the client alone.
The check is deliberately cheap and deliberately first. It costs a few lookups, and a wrong chase costs a client's patience. Putting the cheap, reversible step in front of the expensive, irreversible one is the whole design.
Five Outcomes, Not One: Chase, Hold, No Action, Contact Alternative, Escalate
Once the agent has read the threads and the engagement signals, it must choose exactly one of five outcomes. A chaser that can only say yes or no will always drift towards yes, which is how timers behave. Five outcomes give it somewhere sensible to go when yes is wrong.
Chase. No active thread, quotes delivered, and the engagement signals suggest the client has either not engaged or has engaged and gone quiet. The agent drafts a follow-up. This is the outcome a timer would have produced every time, and here it is the minority result.
Hold. There is an active thread, or the signals show the client is mid-comparison. The agent writes nothing and records why, so the next check starts from a reasoned position rather than a blank one. Holding is not inaction by neglect; it is a logged decision.
No action. The trip no longer needs a chaser: the client has booked, declined, or the departure date has passed. A timer keeps firing on dead enquiries; the agent closes the question.
Contact alternative. The named client is silent, but another person on the trip is not, or the record holds a second contact such as a PA. Rather than send a third email to an inbox that is not answering, the agent routes the follow-up to the person who is. This is the outcome most manual desks reach last, usually after an awkward silence.
Escalate. The situation needs a human: a high-value sector, a client who has opened every quote several times but never replied, a complaint in the thread, or a chaser count that has run out. The agent hands the trip to a person with the evidence attached instead of sending another email.
Restraint is built into this list. Three of the five outcomes, hold, no action and escalate, send nothing to the client at all. That share is the feature, because each of them removes a message a timer would have sent.
The same decision structure could run on other enquiry-led businesses. The architecture could handle a yacht charter enquiry or a high-value vehicle rental quote, where a client is also comparing several proposals across email and phone, though villiersOS does not operate in those sectors today.
Anatomy of a Chaser: Route, Departure Date, Quote Count and Lowest Price in the Right Currency
When the outcome is chase, the email is not a template with fields filled in. The body is generated from a defined set of facts about the trip, and the agent is told what to include and what to leave out.
The context passed in has seven parts: the origin and destination, named in full as airports rather than as codes alone; the departure date; how many quotes are available; the lowest quote price; the dashboard view count; which chaser this is; and the agent's persona and role on the desk. The output is structured: a subject line in the form "Re: Charter [origin] → [destination]" and two or three short paragraphs.
The currency detail matters more than it looks. The lowest quote is shown in the currency that fits the route: euros for a European sector, dollars for a transatlantic one. A London to Nice client expects to see €10,900, not a sterling conversion that reads as an afterthought. Typical one-way pricing on that route runs from roughly €10,500 to €16,500 on a light or midsize jet, which is the £9,000 to £14,000 range quoted above at about €1.17 to the pound. A transatlantic sector from London Luton (EGGW) to Teterboro (KTEB) on a large-cabin jet such as a Challenger 650 typically starts in the region of $55,000 and rises with aircraft size and season, so dollars are the natural frame there.
The chaser number sets the tone. The first follow-up, sent a few hours after quotes go out, simply confirms delivery and offers help. A later one acknowledges that time has passed. A chaser that does not know it is the third message will repeat the second, and clients notice.
The view count shapes the wording without being quoted back. A client who has viewed the dashboard several times gets a note that offers to talk through the options. A client with no views gets a note that checks the quotes arrived. The agent never says "we can see you have looked at the quotes", because that reads as surveillance rather than service.
A worked example: four quotes for EGLF to LFMN on 18 July, lowest at €10,900, two dashboard views, first chaser. The email names Farnborough and Nice in full, gives the date, says four options are available from €10,900, and asks whether any of them should be adjusted. It runs to two short paragraphs and contains no urgency, because there is none to manufacture.
Send Windows, Sign-Off and the Human Review That Keeps a Follow-Up From Becoming Spam
A good decision can still be undone by a badly timed send. Autonomous outbound on the Villiers desk goes out in a 11:00 to 16:00 UK window, on the next business day, never the same day the draft is written. A chaser composed at 17:40 on a Friday goes out on Monday, not at 17:41.
The next-day rule is there because it gives the hold logic one more chance to be right. In the intervening hours the client may ring, reply, or open the dashboard, and the check runs again before sending. A follow-up that is no longer needed is cancelled rather than delivered.
Before any autonomous follow-up is sent, the draft is posted as a comment on the task first, and a human signs it off. This is a deliberate human review step, and it is the safeguard that stops a plausible but wrong email reaching a client. The agent proposes; a person approves.
Copy is policed as well. The register is plain and calm: no pressure, no countdowns, no phrases such as "last chance" or "act now", and no marketing emoji. A copy guard in the sending path rejects drafts that use that language, so the restraint does not depend on anyone remembering it on a busy afternoon.
The sign-off is a name and role on the desk, with no flourish. The chaser reads like a note from a broker who has looked at the trip, because the decision layer has in fact looked at it.
To apply this on your own desk, whether or not you use villiersOS, put four questions in front of every follow-up and answer them in order. Has the client or anyone on the trip been in contact since the quotes went out, by email or phone? Have the dashboard views and quote count suggested the client is still comparing? Is there another person on the trip who is answering? And if none of those explain the silence, is this chaser number one, two or three, and does a human need to take over? Only the last answer should ever produce an email.
