Outcall Dispatch: Turnaround Time Is the Whole Business

A venue's constraint is seats. An outcall agency's constraint is minutes in a car.

That single difference changes every number that matters. A club can be full and still sell more by raising a price. An outcall operation is capped by how many completed jobs fit between the first booking and the last drop-off, and every minute a driver spends empty is capacity that cannot be recovered.

Most agencies run dispatch on a group chat and a whiteboard, and most of them have no idea what their turnaround actually is. This is the operating math, in the terms dispatchers use.

The unit of work

Everything is built on one object: a run. A run has a pickup, a destination, a start, a finish, and a driver. Everything worth measuring is either a property of a run or a gap between two of them.

Term	What it means
Run	one complete job, pickup through drop-off
Dead leg	a driving segment with nobody being delivered
Turnaround	finish of one run to start of the next, same provider
Standby	on shift, no run assigned
Callout	the booking request itself, before assignment

The gaps — dead legs, turnaround, standby — are where the business is won or lost, and they are exactly what nobody records.

Turnaround: the number that caps revenue

Formula.

Turnaround = start of next run − finish of previous run (same provider)
Runs per shift = shift length ÷ (average run duration + average turnaround)

Why it caps revenue. If a run takes ninety minutes and turnaround averages fifty, an eight-hour shift fits three and a half runs. Cut turnaround to twenty-five and the same shift fits four and a quarter. That is a twenty percent revenue increase with no additional bookings, no additional staff, and no price change.

The decision it drives: everything about how you assign. Turnaround is dominated by three things — geographic sequencing of jobs, how quickly the next booking is confirmed, and whether the driver is told where to head before the current run ends. All three are dispatch decisions, not driver behaviour.

Where turnaround actually goes

Broken into its parts, so you know which one to attack:

Turnaround = travel back or onward
           + waiting for the next assignment
           + confirmation delay with the client
           + provider's own break

The second and third are the ones dispatch controls, and in most operations they are the largest. A driver sitting in a car park waiting for the office to decide is the most expensive idle minute in the business, and it does not appear in any report.

Dead legs: the distance nobody is paying for

Formula.

Dead leg ratio = empty distance ÷ total distance
Loaded ratio   = 1 − dead leg ratio

The decision it drives: how you sequence the board. Sequencing by booking time — first come, first assigned — is the default everywhere and is almost always wrong. Sequencing by geography, with booking time as a constraint rather than the ordering key, is what pulls the dead leg ratio down.

Why it goes unmeasured: it requires knowing where the car was, not just when the job ran. If you record pickup and destination on every run, you can compute it approximately without any tracking hardware, and approximate is enough to tell whether your board is sequenced or just queued.

Driver utilization: three different numbers people call the same thing

This is where most agencies confuse themselves, because "utilization" means at least three things:

Time utilization  = minutes on a run ÷ minutes on shift
Loaded utilization = loaded minutes ÷ minutes on shift
Job utilization   = runs completed ÷ runs assigned

They diverge, and the gap is informative. A driver with high time utilization and low loaded utilization is doing a lot of driving and not much delivering — a sequencing problem. High loaded utilization with low job utilization means jobs are being assigned and then falling through — a booking or confirmation problem, not a driving one.

Pick one as the headline and state which. The most common reporting error in dispatch is quoting one of these and having the reader assume another.

Callout to confirmation: the front of the funnel

Formula.

Confirmation rate = bookings confirmed ÷ callouts received
Time to confirm   = median minutes from callout to confirmed assignment
Fill rate         = callouts filled ÷ callouts received

What time-to-confirm decides: whether you keep the booking at all. In this business the client is frequently contacting more than one agency, and the second-fastest confirmation often loses the job outright. Every minute here is a conversion problem, not an administrative one.

What fill rate decides: roster size, by hour. A fill rate that collapses in a specific two-hour window is not a general staffing shortage; it is a schedule that does not match demand, and the fix is cheaper than hiring.

The callout log nobody keeps

Most agencies record bookings. Almost none record callouts that did not become bookings — and that is the record which tells you what you are turning away. Three fields: time, what was requested, why it did not fill. A month of that data usually reorders the roster.

Provider retention: the same lesson as everywhere

Formula.

Shift-weighted retention = providers working ≥ N shifts this period
                           who also worked last period
                           ÷ providers who worked last period

The decision it drives: recruitment timing. It leads revenue by roughly a booking cycle, and rebuilding a roster is slower than losing one.

The dispatch-specific driver of it: turnaround. A provider whose earnings are capped by dead time will find an operation where they are not, and the operational number and the retention number are therefore the same number viewed from two ends.

The dispatch board itself

Whether it is a whiteboard or a screen, the board has to answer four questions at a glance, or dispatchers will keep the real state in their heads:

1. Who is free right now, and where are they? Not "who is on shift" — where they physically are.
2. What is unassigned, and how long has it been? With the time visible, because the delay is the loss.
3. What is running, and when does it end? So the next job can be sequenced before the current one finishes.
4. What is confirmed for later tonight? So the sequence is planned rather than reacted to.

A board that shows only the third of these — which is what a status list gives you — makes every assignment a reaction. Most software in this space is a status list.

The dispatch board we built

Almost everything sold to this industry is a venue product with a dispatch tab bolted on. tasteck was built the other way round: outcall dispatch is one of the parts we built first, because we were running it.

The board answers all four questions

Go back to the four questions a board has to answer at a glance. Ours answers them:

Who is free and where — a live dispatch screen showing provider status and driver status together, not a list of who is rostered. What is running and when it ends — so the next job is sequenced before the current one finishes, which is the largest single cut to turnaround available to you. What is confirmed for later — the night as a plan rather than a reaction. What is unassigned — with the clock visible, because the delay is the loss.

Drivers update arrival and return from their own phone, so the board is current without the office phoning anyone. And a callout arriving on the phone becomes a booking record without retyping, because the reservation screen is wired to inbound calls — which is where time-to-confirm, the number that decides whether you keep the job at all, is actually won.

Then it answers the question dispatch software never does

Turnaround caps your revenue. But the other half of the business is what fills the board in the first place, and that is where we go somewhere no dispatch product goes:

tasteck outputs the maximum you can spend on each advertising channel next month, as an amount in your currency.

Every listing, every directory, every ad account you pay — it takes the lifetime value of the clients that source actually delivered, applies your target margin, and gives you the ceiling. Nothing else in this category produces that figure. A dispatch board tells you how efficiently you served the demand. This tells you what the demand is worth buying.

Ask it from ChatGPT

tasteck connects to ChatGPT over MCP: ask your numbers as a question and get the answer in the chat — runs by driver last week, fill rate by hour, which sources are producing repeat clients. No dashboard, no report to locate.

Measured against every vendor listed on Japan's principal nightlife-industry directory, this is the first implementation of it in the category. The same interface is callable from anywhere, so it is not tied to one assistant.

Built by operators

Every competitor here is an IT firm's model of this business. tasteck was built by people who ran it for sixteen years and grew from ¥200 million to ¥1.2 billion a year, six-fold, by attacking it with systems. The reason the dispatch screen shows provider and driver state on one surface is that this is what it took to run the nights — not what looked good in a wireframe.

Multi-language is built in, the whole operating surface, with your language set put in place during onboarding.

What it does not do

It does not compute dead leg ratio, sequence geographically, or log callouts that did not fill. The first two need route data we do not yet capture; the third is a gap we can see. If wide-geography routing is your single bottleneck, that is worth knowing now.

Where to start

1. Record the finish time of every run, not just the start. Turnaround is uncomputable without it, and most operations record only starts.
2. Log callouts that did not fill. Three fields. It is the cheapest data in this list and usually the most surprising.
3. Measure time-to-confirm for one week. If it is above a few minutes, that is likely the largest recoverable loss in the business.
4. Sequence one night geographically instead of by booking order and compare runs completed. One night is enough to see whether it moves.
5. Tell drivers their next destination before the current run ends. No system required. It is frequently the single largest cut to turnaround available.

Only item four benefits meaningfully from software. The others are a decision and a column.

Find out what your demand is worth buying

The ceiling calculation is open on our site, nothing to sign up for: Ad budget calculator. Put in what you paid each listing or channel last month and what those clients spent. It gives you the ceiling per source, as an amount. Nothing is transmitted anywhere — use it and close the tab.

That is the figure computed by hand, once. The full system produces it every month out of your own bookings, on top of the dispatch board above.

From $34 a month for up to two venues. Thirty days free on every plan, cancel any time. → Pricing

---

On benchmarks. No target ranges are quoted here for turnaround, dead leg ratio, utilization, or fill rate. We do not have a dataset broad enough to publish them honestly — these vary enormously with city density, licensing, and how far apart the jobs are. Measure your own for four weeks; that band is the only benchmark that describes your operation.