What PoolBoss Says
Fire a pool service customer when the account costs more than it pays and a price increase has already been declined. Run four tests - time on site against the price, callback frequency, how reliably they pay, and how the customer treats your technician - and act when two or more of them fail.
Every route has one. The stop your tech schedules last because it always runs long, the owner who texts on Sunday about a spa he keeps refilling with a garden hose, the pool still paying the same $115 it paid in 2021. You already know which account it is. What you do not know is whether dropping it is a business decision or a tantrum.
It is a business decision when the arithmetic says so. The four tests below turn a feeling into a number, the price increase gives the customer a fair chance to fix it, and the exit is a short, unemotional sequence that keeps your name clean in a town where pool owners talk to each other.
At a glance
Key takeaways
- Run four tests before dropping anyone - time against price, unpaid callbacks, how reliably they pay, and how they treat your tech - and act only when two fail.
- Quote the real price first. A customer who declines a fair increase has ended the relationship for you.
- Close the whole gap when you reprice: a $115 stop that burns $80 of capacity is not fixed at $130.
- Check geography and an old price before you blame the person - most money-losers are routing or pricing problems.
- A technician who does not want to go back to a house is the one test that ends an account by itself.
- Give 30 days' notice in writing with a specific last service date, and attach no list of reasons.
- Keep any unpaid balance on its own track; bundling money into the exit is what turns it into a review.
When should I fire a pool service customer?
When two of the four tests below fail and a price increase has already been declined. One failing test is a problem to fix, not a reason to end a relationship - a stop that runs long might just need a different day, and a slow payer might just need autopay. Two failing at once is a pattern, and patterns do not resolve themselves on a recurring route. A weekly account is on your schedule about 50 times a year, so a problem you tolerate in March is one you will still be carrying in November.
The tests are deliberately narrow, because the reasons operators give for dropping an account are usually broader and softer than the reasons that hold up afterward. "He is difficult" is not a test. "He costs me 55 minutes for a stop priced at 25" is. Write the answer to each test down against a specific account, because the moment the numbers are on paper the decision usually makes itself - and if it does not, you have learned something useful: the account is fine and the irritation is yours to manage.
All four tests run on information you already generate every week - when the tech arrived and left, what was dosed, what was invoiced, what was paid. The work is putting it beside the price. Knowing what each account actually costs you to serve is what separates an operator who prunes his route deliberately from one who carries four money-losers for a decade because none of them ever did anything bad enough to notice.
The four tests that separate a bad account from a bad mood
Each test has a threshold you can check and a fix to try before the account goes. Run all four against the same account before deciding, because the pattern matters more than any single result - and three of the four have a cheaper answer than firing anybody.
| Test | What to measure | Where the line sits | Try this first |
|---|---|---|---|
| Time against price | Minutes on site plus the detour to reach it, against the stop's price | The time costs more than the stop pays, or it runs 2x your route average | Reprice it to the time it takes, or move it to a day you are already nearby |
| Callback frequency | Return trips you are not paid for, per month | More than one unpaid callback a month, three months running | Fix the cause, or price callouts separately from the monthly rate |
| How reliably they pay | Days past due, and how many cycles in a row | Late two or more cycles running, or a balance past 30 days | Move them to autopay, or to prepayment for the month ahead |
| How they treat your tech | Cancellations at the gate, abuse, unsafe access | Any safety issue, or a tech who asks not to go back | One direct conversation, in writing |
- Two failures is the trigger. One is a maintenance job on the account, not grounds to end it.
- The safety row is the exception: a technician who does not want to go back to a house ends the account on its own, and no price fixes it.
- Measure over three months, not one bad week. A single July of callbacks on a pool that has been fine for four years is a repair problem, not a customer problem.
- Run the tests on your best account too, once. Operators are routinely wrong about which stop is actually costing them.
Raise the price before you drop the account
The increase is the honest version of the decision, and it settles the account either way. Price the stop at what it actually takes rather than what it has historically paid, quote that number, and let the customer choose. If they accept, the account stops losing money and you keep the revenue. If they decline, they have ended the relationship themselves - which is an easier conversation than you ending it, and a much easier one to describe later if anyone asks.
Getting the number right matters more than the delivery. The increase has to close the whole gap rather than split the difference, because a $115 stop that consumes $80 of capacity does not become healthy at $130. Work out the real cost of the visit first, add your normal margin, then quote it. The mechanics of raising the price on an existing account - notice, timing, and how to word it - are worth doing properly, because the same letter goes to customers you fully intend to keep.
Two numbers make the quote defensible rather than guessed: what the pool costs you in chemicals, and what the route earns against that cost. Per-pool chemical cost and per-route profitability come out of the reports PoolBoss builds from the visits you already log, on every plan including the free one. What no software will do is tell you the account is bad. It has no idea the owner meets your tech at the gate every week to explain the job. It gives you the cost side; the judgment stays yours.
A bad account is usually a routing problem or a pricing problem wearing a personality
Before deciding a customer is the problem, check whether geography or an old price is. The two most common money-losers on a route are not difficult people at all: the stop that sits well off the loop and pays the same as everyone else, and the underpriced legacy account that has not moved while wages and chemicals rose around it. Both leave you irritated every week, which is why both get mistaken for bad customers, and neither one requires firing anybody.
Take a solo operator running 51 pools around Katy and Cypress. One Cypress account pays $115 a month, sits 14 minutes off the loop, and averages two callbacks a month over a spa the owner keeps refilling from a garden hose. Time on site plus the detour runs about 55 minutes against a 25-minute route average, so the stop burns roughly $80 of capacity to bring in $115 - before the callbacks, which are free. He prices it properly at $165 and offers it. The customer declines. The stop comes off with 30 days' notice, and the drive time it was eating absorbs two Katy pools he had been turning down for lack of room.
That last clause is the part operators miss. Dropping an account does not simply cost you its revenue; it trades that revenue for capacity, and on a full route capacity is worth more than the worst stop on it. Whether the trade is a good one depends on the per-route numbers behind the call. With a waiting list, releasing a money-loser is close to free. With three empty afternoons, fix the price or the routing first and keep the account.
How to end it without a review problem
Do it in writing, once, without a list of grievances. The sequence is short: give notice with a specific last service date, send the final invoice for work already done, arrange the return of keys, remotes or any equipment of yours, and stop the recurring visit so nobody shows up by accident. Thirty days is the normal courtesy on a month-to-month residential account, and it is long enough for the customer to find someone else - which is the single biggest factor in whether this ends quietly.
Say less than you want to. "We are no longer able to service your pool after October 15" needs no justification attached, and every reason you add is something that can be argued with. Operators get into trouble writing the honest paragraph - the callbacks, the spa, the Sunday texts - and handing the customer a bill of particulars to dispute in public. If they ask why, "the route has changed and your pool no longer fits it" is true in almost every case, because it is nearly always the routing or the price that made the account unworkable.
Money owed is a separate track from the exit and should not be folded into it. Send the final invoice on its normal terms and follow your normal reminder sequence; the standing policy for an account that keeps paying late does not change because the service is ending. Bundling the two turns a clean ending into an argument about money, and an argument about money is the version that gets written up.
Then check the account against the ones you are working to keep. A customer failing two tests after a declined increase looks nothing like one who is simply annoyed about something fixable, and the tests are what tell them apart. Most of your route belongs in the second group - the accounts worth working to keep are the overwhelming majority, and the whole point of having a test is that firing anyone stays rare.
FAQ
Frequently asked questions
How much notice should I give a customer I am dropping?
Thirty days is the standard on a month-to-month residential account, because it covers a full billing cycle and gives the customer time to find a replacement. Check three things before you pick the date: what your own service agreement says, whether the customer has prepaid for a period running past your date, and whether you are mid-cycle on a monthly rate. If they have prepaid, either service through the paid period or refund the unused portion - keeping money for visits you will not make is how a quiet exit becomes a complaint. Put a specific last service date in writing rather than "the end of next month", so there is no argument later about which Tuesday was the last one. The exception is a safety issue or abuse of your technician: you do not owe 30 days for that, and you should stop immediately and say so in writing.
Should I tell the customer the real reason?
Give a short, true, non-specific reason and stop there. "The route has changed and your pool no longer fits it" is accurate in most cases, because it is almost always the drive time or the price that made the account unworkable. The detailed version - the callbacks, the Sunday texts, the spa - reads as a list of accusations, and every item on it is something the customer can dispute in public. Two exceptions are worth making. If the reason is price and you would genuinely keep the account at the right number, say so plainly and quote it, because that is not a reason to leave, it is an offer. And if the reason is how your technician was treated, say that directly and without softening it into a routing excuse - it is the one case where the customer should hear it straight.
What do I do if they owe me money when I drop them?
Invoice it and collect it on your normal terms. The balance does not become more or less collectible because the service is ending, so send the final invoice for work already completed with the due date you always use and follow your usual reminder sequence. What you should not do is make payment a condition of the exit, or let the unpaid balance become the stated reason for it - that converts a clean ending into an argument about money, and those are the endings that get written up. If the balance is small, decide quickly whether it is worth pursuing and then be finished with it either way; the worst outcome is a debt you neither collect nor close, attached to a customer who is now talking about you. Stop the recurring invoice as well as the recurring visit, so the account does not keep generating charges after the last service date, and keep the visit and invoice records in case the balance is ever disputed.
Will firing a customer hurt my online reviews?
Sometimes, and how you exit matters far more than the fact of it. The reviews that follow a released account are rarely about the dropping itself - they are about surprise, a disputed final bill, or a reason the customer found insulting. A dated notice with a specific last service date, a final invoice for work actually done, and no explanation beyond a routing change removes all three. If a negative review does land, respond once, briefly and factually, without relitigating: confirm the service dates, note the notice you gave, and offer to resolve any billing question directly. Do not answer the emotional content. Prospective customers reading it are judging how you respond at least as much as the complaint itself, and a calm two-sentence reply under an angry paragraph usually reads in your favor. One review from an account you deliberately released is a cheaper problem than carrying that account for another three years.
Should I refer them to another pool company?
Only if you would genuinely stand behind the referral. Handing a difficult account to an operator you know is a good way to damage that relationship, and the pool trade in any given metro is small enough that it gets back to you. A neutral version works better: tell the customer several companies serve the area and let them choose, or point them at a supply house that keeps a list. There is one case where a referral is genuinely right - when the account is a poor fit for you specifically rather than a bad account in general. A commercial spa that needs a certified operator, a house on the far side of town, a pool that needs twice-weekly service you cannot staff: those are real referrals, and the operator who takes them is getting work they actually want. Say which case it is when you pass it along.
What if it is a commercial account under contract?
Read the contract before you do anything. Commercial agreements usually specify a termination clause, a notice period, and sometimes a cure period that requires you to give the other side a chance to fix the problem first - a different process from ending a residential account, and skipping the stated steps can put you in breach even when the underlying reason is sound. Document the problem while you are still servicing: dated visit records, photographs, and written notice of each issue and what you asked for. If there is a cure period, use it properly and in writing. Check what happens to any prepaid amount, and whether the agreement auto-renews on a date you would need to act before. This is not legal advice, and a commercial contract worth real money is worth an hour of a lawyer's time before you send anything.
Can I drop more than one account at the same time?
You can, but stagger them unless the accounts are actively costing you money right now. Dropping four stops in one week takes the revenue out immediately while replacements take weeks to land, and on a route with spare capacity that is a self-inflicted cash flow problem. The safer sequence is one at a time, worst first, moving to the next only once the freed capacity has been filled - either by a new account or by tightening the route around the gap. There is one exception: when several money-losers sit clustered in the same outlying area, releasing them together is what actually recovers the drive time, because dropping one of six distant stops saves you almost nothing. Treat that cluster as a single decision, and expect a larger revenue dip in exchange for a larger operational gain.
