What PoolBoss Says
Dispatch off the route, not around it. Keep the recurring stops fixed, then place each unplanned job with the technician whose remaining stops sit closest to it, working from the day's stop order rather than from who answers the phone. Protect the last hour of every route as the slot that absorbs the day's surprises.
The call always comes at the wrong time. It is 10:40 on a Tuesday, three trucks are already out, every hour of the day is spoken for, and a customer is standing next to a pump that is spraying water across their deck. Nobody is free. Somebody still has to go.
That decision is dispatch, and it is a different job from the two that come before it. Scheduling put the stops on days and people last month. Routing put them in driving order this morning. Dispatch is what happens when reality lands on top of both while the trucks are moving. Do it by instinct and you pay for it in second trips, late stops, and the repair revenue you had no room to take.
At a glance
Key takeaways
- Dispatch off the route, not around it - the recurring stops stay fixed and the unplanned job finds a home inside them.
- Run the filters in order: geography narrows the field, skill vetoes anyone who cannot finish, and slack breaks the tie.
- Sending the closest tech who cannot do the work costs two trips on a job that priced for one.
- A 15-minute detour is 30 minutes of real cost - worth it for a $350 repair, not for a $95 filter clean.
- Hold back 45-60 minutes on every truck; that float is the only slot unplanned repair revenue can land in.
- A green-pool recovery needs 60-90 minutes and its own slot, not a squeeze between two cleaning stops.
- Send five things with every dispatched job: address and gate code, what the customer reported, what to bring, which stops moved, and who calls the customer.
How do I dispatch pool service jobs efficiently?
Dispatch off the route rather than around it. The recurring stops are already placed, already promised, and already billed, so the unplanned job is the thing that has to find a home inside them. The question is never who is free, because at 2-6 trucks nobody is free. It is which technician's remaining stops sit closest to the new work, whether that technician can actually do it, and what moving their last few stops costs you.
Most operators are already running three layers without naming them, and naming them is what makes a bad morning fixable. Scheduling decides, weeks ahead, which pools get serviced on which day by which technician. Routing decides the order those stops get driven. Dispatch is the live layer on top of both: the 10:40am call, the equipment failure, the gate that will not open, the tech who finishes 40 minutes early. Scheduling and routing get done sitting down. Dispatch happens while four trucks are moving, which is exactly why it needs a standing rule instead of a fresh debate every time the phone rings.
The rule matters more than the tooling at this size, because the dispatcher is usually the owner and is often also driving a route. Solid pool service routing and scheduling earns its keep by keeping the recurring half stable, so the exception is the only thing you have to think about. In PoolBoss an unplanned repair is a work order in its own right - it carries the customer, the pool, an assigned technician, a date, and a price - and it lands on the same shared week grid as the route visits, so you can see that Thursday is already carrying 16 stops before you drop a seventeenth onto it.
What none of this requires is live vehicle tracking or an algorithm that reassigns trucks on its own. Most independent operators do not have either, and at four trucks the honest version is better anyway: you know your roads, you know which tech can pull a pump seal, and the whole decision takes about 90 seconds once you make it in a fixed order.
Not every job goes to whoever answers the phone
Sort the work by type before you sort it by person. Job type answers three separate questions that operators tend to collapse into one: how urgent it genuinely is, who is qualified to do it, and what it does to the route it lands on. A leaking pump and a new-customer first clean are both unplanned work, and almost nothing about how you place them is the same.
| Job type | Who should take it | How urgent | What it does to the route |
|---|---|---|---|
| Emergency leak or pump failure | Whoever carries repair tools and parts, even if they are not closest | Same day | Displaces 2-3 stops that slide to another day |
| Non-urgent equipment repair | A repair-capable tech, booked as its own job | Within the week | None, if it gets a real slot instead of a squeeze |
| Green-pool recovery | An experienced tech - this needs 60-90 minutes, not 20 | Within 24-48 hours | Eats a full stop slot and usually a return visit |
| Return trip for a missed stop | The tech whose route it was | Next day, at the start or end of a route | Adds one stop to a day already driven |
| New-customer first clean | Any tech, but it is the longest of the five | Whenever it fits the week | Plan it in - never insert it same day |
- Only the first row is a true same-day emergency. Treating rows 2, 3 and 5 as emergencies is how operators end up running a whole business at emergency pace.
- The skill question is a hard filter, not a preference. Sending a tech who cannot finish the job means a second trip, and a second trip costs the drive twice while earning the job once.
- A green-pool recovery is not a stop, it is most of an afternoon. Book it as its own job with its own time, or it silently pushes six other customers late.
Place the job by geography, then by skill, then by slack
Run the three filters in that order every time. Geography narrows the field to the trucks that could plausibly get there. Skill is a veto that removes anyone who cannot finish the work. Slack breaks whatever tie is left. Reverse the order and you get the most common dispatch mistake in the trade: the nearest truck takes a job it cannot complete, and you pay for two trips on a job that priced for one.
The drive-time arithmetic is what makes this concrete. A 15-minute detour is really 30 minutes once you drive back to where the route left off, and dropping that into an afternoon with 6 stops left pushes every one of those customers half an hour later. If the job at the end of the detour is a $95 filter clean, it did not pay for the disruption. If it is a $350 pump repair, it did. Same detour, opposite answer, and the only variable is what the job is worth - which is why it is worth knowing what a detour actually costs in drive time before the phone rings rather than during the call.
Here is the whole decision in one real morning. A four-truck operator working Katy and west Houston takes a call at 10:40am: a pump seal is leaking at a Cinco Ranch property and water is running across the deck. Tech A is on the far side of Katy with 6 stops to go. Tech B is two stops away from Cinco Ranch, which makes him the obvious answer, except he is the newest hire and has never pulled a pump. Tech C carries the repair kit, is finishing a green-pool recovery in Fulshear 19 minutes out, and has two stops left that can slide to Thursday without anyone noticing. The job goes to Tech C. The two slipped stops move to Thursday's route, and the Cinco Ranch customer gets a two-hour window before anyone drives anywhere. Sending Tech B because he was closest would have bought a look, a phone call, and a second trip - which is the mistake that costs the afternoon rather than the 19 minutes.
When the honest answer is that nobody should take it today, say so on the first call. A customer told at 10:45 that a tech will be there at 8am tomorrow will usually take it. The same customer, told at 4:30pm that nobody made it, will not, and that is the version that ends up as a review.
Protect the last hour of every route
Leave the last 45-60 minutes of each route unassigned on purpose. That single habit is what turns dispatch from a daily crisis into a placement decision, because a route packed to the final minute has no capacity for the work that actually pays.
The math is not subtle. If each of four trucks runs 8 stops with zero float, an emergency call has exactly two possible outcomes: you refuse the work, or you displace a cleaning stop you have already promised and billed for. Hold back an hour a truck and you have four hours of daily capacity that can absorb a repair, a green-pool recovery, or a return trip without touching anyone's recurring service. Across a five-day week that is roughly 20 hours of slack, and on a route that never uses it, the techs finish early - which is not a loss.
That buffer is where the margin lives, too. Recurring cleaning is the base of a pool business, but repairs and equipment work carry the better rate per hour, and they arrive unplanned by definition. An operator with no float cannot say yes to them. Booking every hour of every truck looks like efficiency on a spreadsheet and behaves like a ceiling in practice.
This is also the practical answer to whether pool work belongs on routes or on individual jobs, which is a question most operators wrestle with once they pass two trucks. The workable arrangement is both: recurring cleaning lives on routes, and unplanned work gets booked as its own job into the float. If you have not settled that question yet, the tradeoff between routes versus individual jobs is worth reading first, because dispatch only works cleanly once the recurring half is stable.
The handoff is the part that breaks
A job that is dispatched but not communicated is a job that does not happen. This is where most dispatch actually fails - not in the decision about who goes, but in the 30 seconds afterward that nobody spends. The tech is driving, the office moved on to the next call, and the work sits on a screen someone will look at tonight.
Five things travel with every dispatched job, and a job missing any of them tends to come back: the address and the gate code, what the customer actually reported in their own words, what to bring, which of the tech's own stops are being displaced and where those went, and who is calling the customer with the window. That last one is the item that gets dropped most often, usually because both people assume the other has it.
None of it is automatic, and it should not be described as if it were. Software can put the job on the right technician's day, hold the address and the gate code where a phone can reach them, and record that the stops you slipped moved to Thursday. It does not make the call to the customer, and it does not tell the tech that the job is urgent in a tone they will hear. At four trucks, the text message is still the dispatch system; the software is the record that keeps the text message honest.
One-off dispatch and losing a whole route are different problems with different answers, and it is worth knowing which one you are in. Placing a single repair into three working routes is the exception handling described here. Covering 40 stops because someone is out sick is a reassignment problem - how to reassign a whole route when a tech is out is its own decision, and trying to solve it one job at a time is how a Tuesday gets away from you.
FAQ
Frequently asked questions
Should my route techs do repairs, or do I need a separate repair tech?
Route techs handle repairs until roughly the third truck, and after that the economics usually favor splitting the work. A cleaning route runs on rhythm - 8 to 12 stops a day at 20-30 minutes each - and a repair dropped into that rhythm breaks it for every customer downstream. One person doing both is fine at one or two trucks because there is nobody else to send. Past that, a dedicated repair tech who carries parts and takes the unplanned work protects the cleaning routes from constant disruption and usually bills at a better rate per hour. The halfway step most operators take first is designating one route tech as the repair-capable one, keeping their route slightly lighter, and sending all equipment work to them. That works well up to about four trucks. What does not work is assuming every tech can do everything, because the day you dispatch on that assumption is the day you pay for a second trip.
How do I price a same-day emergency pool call?
Price the disruption separately from the work. Most operators charge a trip or diagnostic fee of $75-$150 for a same-day call, then bill parts and labor on top at their normal repair rate. The fee is not opportunistic - it covers the drive, the stops that got pushed, and the fact that you rearranged three people's afternoons. Say the number on the phone before anyone drives, along with whether it applies if the fix turns out to be trivial. The conversation that goes badly is always the one where the customer expected the visit to be free because they are already a monthly customer. It also helps to have a stated policy for existing accounts versus new callers: many operators waive or reduce the trip fee for customers already on a recurring route and charge it in full for a one-off caller, which is defensible and easy to explain.
What do I do when a repair call comes in and every tech is already booked?
Decide whether it is genuinely an emergency, then either displace a stop or book it first thing tomorrow - and tell the customer which within the hour. A true emergency is water leaving the pool, an electrical problem, or a safety issue; those justify pushing 2-3 cleaning stops to another day. Almost everything else is a next-morning job, and customers accept next-morning far better than they accept silence followed by a no-show. If you do displace stops, move them somewhere real rather than leaving them floating, and let those customers know their service moved - a cleaning customer who quietly gets skipped is a cancellation in three months. The pattern worth watching is frequency: if every week produces a call that nobody can take, the routes are packed too tight, and the fix is float rather than heroics.
How much buffer should I leave in a pool route for unplanned work?
Roughly 45-60 minutes per truck per day, which works out to about 10-15% of the working day. On an 8-stop route that is the equivalent of two stops you deliberately do not book. It feels expensive right up until the first week you use it, because that hour is what lets you say yes to a repair without breaking a promise to a cleaning customer. Operators who run at 100% capacity are not more efficient; they have simply moved the cost from visible idle time to invisible late stops and refused work. If your route genuinely never needs the buffer, the techs finish an hour early and you have room to add accounts - which is a good problem and a much easier one to fix than a route with no give in it. Seasonality matters too: hold more float in peak summer, when equipment fails and pools go green fastest.
How do I tell a customer their regular visit is running late?
Message them before they notice, and give a window rather than an apology. A short text that says the tech is running behind and will be there between 2 and 4 does almost all the work; customers are far more tolerant of a late visit than of an unexplained one. The failure mode is waiting until the end of the day to see whether you will make it, because by then the customer has already decided you did not show. If the stop is going to move to another day entirely, say that explicitly instead of leaving it vague, and name the day. For customers who are never home and do not care when service happens, this matters less - but you should know which of your accounts those are rather than assuming. Anyone with a gate code, a dog, or a pool-day preference needs the message.
Should a return trip for a missed stop count as a new job?
Track it as its own visit, but do not bill it as an extra one if the miss was yours. The record matters: a return trip is real drive time and real labor, and if you never log it you will never see that one account is generating three return trips a quarter because of a locked gate. The billing question is separate and turns on cause. If the tech skipped the pool because the route ran long, that is your cost to absorb. If the gate was locked, the dog was out, or nobody left access after being asked twice, most operators bill it - and the ones who do not usually end up with a customer who never fixes the access problem. Whichever you choose, decide it as a policy rather than case by case, and put it in the service agreement so the first conversation about it is not happening at the point you are trying to charge for it.
At how many technicians do I need a real dispatch process instead of texting?
Texting stops working at about three technicians, and it fails badly at five. With two trucks the owner holds the whole day in their head and a text is genuinely enough. At three the failure starts: someone gets a job by text, the office does not record it, the customer calls about it, and nobody can say what was promised. At five it becomes the main source of lost work. The process that replaces it does not need to be elaborate - one shared place where every unplanned job is written down with a customer, a pool, an assigned tech and a date, so the answer to what was promised does not depend on scrolling a phone. The text message stays; it is the fastest way to reach a driving technician. What changes is that the text becomes the notification and the shared record becomes the truth.
