HVAC
HVAC warranty reserve and callback cost management field guide
The callbacks land after you booked the profit: what the warranty obligation costs, what a callback really runs, sizing a reserve from your own history, and killing the root cause so the same call does not come back.
Direct answer
Warranty cost management is pricing, reserving for, and reducing the callbacks and rework that land after a job closes. Every job carries a warranty obligation, and honoring it costs real money after you already booked the profit. Set aside a reserve from your own history and kill the root causes. This is general education, not accounting or legal advice.
Key takeaways
- Warranty cost management means pricing, reserving for, and reducing the callbacks and rework that land after a job closes and the profit is booked.
- A callback costs far more than the part: add tech labor, round-trip drive, fuel, truck wear, office time, and the billable work not done.
- Size a warranty reserve from your own callback history as a percentage of revenue, broken down by job type. There is no universal correct percentage.
- Log every callback with the originating job, cost, cause, crew, and a workmanship-or-equipment tag so history can size reserves and rank causes.
- Tag each callback workmanship or equipment: workmanship draws your reserve, equipment defects route to a manufacturer claim for recovery.
What warranty cost management is, and why the callbacks eat the profit you already booked
Warranty cost management is the work of knowing what your callbacks and warranty obligations actually cost, setting money aside to cover them, honoring them efficiently, and driving down the root causes so the same call does not come back. It is not the warranty claim paperwork, and it is not the legal language in the contract. It is the money side: the part of every job that gets spent after the job is closed.
Here is the trap. You finish the install, you invoice it, and you book the profit. The job is done on paper. Then three months later the condensate line plugs, or the start capacitor that was marginal fails on the first hot day, or the customer calls because a zone never balanced right. A tech rolls a truck, spends two hours, burns a part, and bills nobody. That cost is real, and it lands after you already counted the profit on that job. Do that across a year of work and the callbacks quietly carve a slice off a number you thought was banked.
Managing it has four parts that this guide works through in order. Know your true warranty and callback cost. Reserve for it so the cost is matched to the job instead of ambushing a future month. Honor the obligation fast and cleanly, because the warranty visit is also a relationship. And most of all, find the root cause and fix it, because the cheapest callback is the one that never happens. Two neighbors to this guide carry the pieces it leans on. The change order guide covers holding the line when a warranty call turns into new paid work. The warranty claim processing guide covers recovering a manufacturer defect from the manufacturer. This guide is the cost and reserve side of the same problem.
One thing before any of it. This is general education, not accounting or legal advice. How you book a reserve, what percentage is right for your shop, and what your contract actually obligates you to are questions for your CPA and your warranty terms, not for a field guide.
Read this first: education, not accounting or legal advice
Everything here is general field education. It is not accounting advice, it is not tax advice, and it is not legal advice. The accounting treatment of a warranty reserve, whether and how you accrue it, the percentage that fits your numbers, and how any of it hits your financial statements or your taxes are decisions for a construction CPA who knows your books. Do not set up a reserve on the strength of a guide.
The same goes for what you owe. What your warranty covers, how long it runs, and what counts as a defect versus normal wear or owner neglect are governed by your contract, your warranty terms, and the law in your state. The examples and any numbers below are illustrative, picked to show how the math behaves, not to tell you what is correct for your shop. A percentage that fits one contractor's failure history is wrong for another.
Treat the two professionals as your backstops. The CPA owns the books and the accrual. The contract and the warranty language own the obligation. This guide helps you ask them the right questions and run the field side well. It does not replace either one.
Why do callbacks eat profit you already booked?
Callbacks eat booked profit because the revenue and the cost land in different months. You recognize the profit when you close the job. The warranty cost shows up later, sometimes much later, and by then it reads as a random expense in a month that has nothing to do with the job that caused it. The P&L looks fine on the job and then bleeds quietly somewhere downstream.
That timing gap is the whole problem. Without a reserve, every callback is a surprise that hits current margin instead of the job it belongs to. A good month gets dragged down by warranty work on jobs you finished last spring, and you cannot see the connection because the cost is not tied back to the source. You end up managing a number you do not understand.
It is a margin issue and a reputation issue at the same time. The margin side is the money you spend honoring the obligation. The reputation side is what the customer remembers about how you handled it. A shop that ignores both treats callbacks as noise. A shop that manages them treats every callback as a cost to measure, a reserve to draw from, and a signal about something upstream that needs fixing. The difference between those two shops shows up in the year-end number.
What does your warranty actually obligate you to?
Your warranty obligation is whatever your contract and your warranty terms say it is, and you need to read them rather than assume. The common shape in this trade is a workmanship warranty on your labor for about one year after substantial completion, sitting alongside the manufacturer's equipment warranty on the parts, which often runs longer. Those are typical patterns, not a rule. Your documents control.
Keep the two clocks separate in your head, because they cover different things and they expire at different times. The workmanship warranty is your promise that the install was done correctly: the brazed joints hold, the system is charged right, the duct is sealed, the wiring is correct. The equipment warranty is the manufacturer's promise that the compressor, the board, the motor, and the coil are free of defects for their term. A failure is one or the other, and which one it is decides who pays.
Watch the difference between a callback period and a warranty of quality, because they are not the same and the distinction has teeth. A one-year callback period is often the window during which you will voluntarily come back and fix defects. The underlying obligation to have built it in a workmanlike manner can outlast that window, and how long it actually runs is a legal question tied to your contract and your state's statutes. That part belongs with a construction attorney, not a field guide. Know what your paperwork says before a dispute forces you to read it for the first time.
What does a callback actually cost?
A callback costs far more than the part, and most shops never add it up. The visible cost is the replacement part and the tech's wage for the hour or two on site. The hidden cost is everything around it: the drive time both ways, the fuel and the wear on the truck, the dispatch and office time to schedule it, and the warranty parts off your own shelf.
The biggest line is the one that never appears on any invoice. It is the billable work that tech did not do because they were on a warranty call instead. Every hour spent fixing a job you already closed is an hour that could have been generating revenue on a job you have not sold yet. That opportunity cost is the real weight of a callback, and it is exactly the number a shop that only counts parts will miss.
Run it loosely to feel the size. Say a callback is two hours of tech time, an hour of round-trip drive, a small part, and the truck. Add the office time to book it. Then add the half-day of billable work that did not happen because the schedule got pulled apart to fit it in. A callback you would have shrugged off as a free part is suddenly a few hundred dollars of real cost, and that is before you count what it did to the customer's opinion of you. This is illustrative, not a figure to quote. The point stands: a callback is never free.
What is a warranty reserve?
A warranty reserve is money you set aside against the warranty cost a job will generate later, so the cost is matched to the job that caused it instead of surprising a future month. The idea is simple even though the accounting is not. When you book the revenue and the profit on a job, you also recognize that the job will probably cost you something in warranty down the road, and you reserve for that expected cost up front.
The reserve, sometimes called a warranty accrual, turns a random future expense into a planned one. Instead of every callback hitting current margin out of nowhere, the callbacks draw down a pool you already funded out of the jobs that created the risk. The good month no longer gets gutted by warranty work on last year's jobs, because last year's jobs already set their share aside.
Whether you should carry a reserve, how it is structured, and how it lands on your financial statements are accounting questions, and they belong with your CPA. There are real rules about how warranty obligations get recognized, and they depend on the kind of warranty and the way you do business. Do not improvise this on the books. The concept is yours to understand. The treatment is the CPA's to set.
How do you size a warranty reserve?
You size a warranty reserve from your own history, not from a number you read somewhere. The honest version is to look at what your warranty and callback work actually cost over the last few years and express it as a percentage of the revenue that produced it. If your real warranty cost has run around some percent of revenue, that history is the starting point for what to reserve going forward. Data, not a guess.
Break it down by job type, because the risk is not uniform. New equipment changeouts, ductwork, controls, light commercial rooftop work, and big custom jobs each carry their own callback profile. A shop that lumps them together reserves too much on the safe work and too little on the work that actually bites. The more your history is sorted by the kind of job, the closer your reserve gets to reality.
Two cautions. First, your history is only as good as your tracking, which is why the tracking section matters before this one can work. If you never logged the callback cost, you have nothing to size from. Second, the percentage is yours and it is illustrative when it appears here. There is no universal correct warranty percentage. The right figure for your shop comes from your numbers and your CPA's read of them, and it moves as your work and your quality change.
The accrual, and where the CPA takes over
The accounting idea behind a reserve is matching: the cost of a job, including the warranty cost it will generate, belongs in the same period as the revenue it earned. Booking the expected warranty cost when you recognize the revenue keeps the job's true profit on the job, rather than letting the cost wander off into a later month. That is the principle. The mechanics are where you stop and call the CPA.
Warranty obligations have specific accounting treatment, and it is not one-size-fits-all. Under the revenue recognition rules, the standard known as ASC 606, the treatment turns on what kind of warranty it is. A warranty that just assures the work was done right is generally handled as an accrued obligation. A warranty that sells the customer an additional service is treated differently, as a separate piece of the deal with its own revenue recognition. Which bucket yours falls into changes the entries, and getting it wrong is the kind of mistake an audit or a tax review finds.
So draw the line clearly. You own understanding why the cost belongs to the job. Your CPA owns how it gets recorded, which standard applies, what the entries are, and how it shows up on the statements and the tax return. Bring them your warranty history and your job types, and let them build the treatment. Do not set up an accrual off a field guide.
Tracking every callback
You cannot manage what you do not measure, and warranty cost is the most-ignored number in most shops. Every callback should be logged with the same few facts: which job it came from, what it cost in parts and labor and time, what caused it, and which crew did the original work. Without those four, you are flying blind on the single largest hidden cost in the business.
The reason most shops do not track it is that a callback feels like nothing in the moment. The tech fixes it and moves on. Nobody writes down that it happened, what it cost, or why. So the cost disappears into general labor and the cause is never known, which means it cannot be sized for the reserve and it cannot be fixed at the source. A callback that is not recorded is a lesson you paid for and threw away.
This is exactly the kind of thing a field tool earns its keep on. In FieldOS, a callback gets logged against the original job with its cost, its cause, and the crew, so the callback rate and the warranty cost build up as real history instead of memory. That record is what feeds everything downstream: the reserve you size, the pattern you find, the crew you retrain, and the manufacturer defect you recover. Capture it at the truck, while the tech still knows what they found, or it does not get captured at all.
Root cause: the highest-value move you have
The highest-value thing you can do with warranty cost is not reserve for it better. It is to make less of it. Reserving covers the cost. Killing the root cause removes it. A callback you prevent costs nothing, generates no unhappy customer, and frees the tech to do billable work. Every other move in this guide manages the symptom. This one cures it.
Root cause means finding the actual reason a callback happened and fixing that, instead of just fixing the unit again. The condensate line that keeps plugging is not a mystery to solve one call at a time. It is a trap detail or a slope the crew is getting wrong on every install, and the fix is the install standard, not the next service call. When you treat callbacks as data instead of nuisances, the patterns show up fast.
The patterns usually point at one of three things: a crew, a detail, or a product. A particular crew's jobs come back more often, which is a training problem. A particular detail fails across crews, which is a process or a standard problem. A particular part or model fails on its own, which is a product problem you route to the manufacturer. Naming which of the three you are looking at tells you where the fix lives. Fix it once at the source and you stop paying for it on every future job.
Reading the callbacks by cause, crew, and type
Once you are tracking callbacks, sort them by cause, by crew, by job type, and by product, and the few causes that drive most of the cost jump out. This is the old Pareto idea and it holds up on warranty work as well as anywhere: a small number of causes generate the bulk of your callbacks. Find that handful and you have found where the money goes.
Sort by cost, not by count, because a frequent cheap callback and a rare expensive one are different problems. Ten plugged drains might cost less than two refrigerant leaks that each pulled a tech for a day. The list ranked by total cost tells you what to fix first, which is not always the thing you hear about most often.
The discipline is to act on the top of the list and ignore the long tail for now. You will never get callbacks to zero, and chasing every one-off is a waste. The two or three causes at the top of a cost-ranked list are where a process change, a training session, or a product switch pays back the most. Fix those, watch the number move, and re-rank.
The repeat callback is a process problem, not bad luck
The same issue showing up twice is not bad luck. It is a process, training, or product problem telling you about itself, and it deserves to be escalated rather than fixed again quietly. A one-off can be anything. A repeat is a pattern, and a pattern has a cause you can name and remove.
When you see the same failure across different jobs, the question is which of the three it is. The same crew, same detail means retrain that crew on that detail. Different crews, same detail means the standard itself is wrong or unclear, and the fix is the install procedure everyone follows. Same part across the board means the product is the problem and the manufacturer needs to hear it.
Escalate the repeat out of the service queue and into whoever owns quality. A callback that is just fixed again, with no one asking why it keeps happening, will keep happening, and you will keep paying for it. The repeat is the cheapest warning you will ever get that something upstream is broken. Spend it.
What is the difference between workmanship and equipment warranty?
Workmanship warranty covers your labor and how you installed it. Equipment warranty covers the manufacturer's parts and whether they were defective. Separating the two on every callback is the difference between eating a cost and recovering it, because they are paid by different parties. Get it wrong and you pay for a defect the manufacturer owed, or you waste a claim on something you broke.
The field test is to ask what actually failed and why. A joint that leaks, a system charged wrong, a duct that was never sealed, a control wired backward: that is workmanship, it is on you, and it draws from your warranty cost and your reserve. A compressor that fails internally, a board that dies, a factory coil that leaks at a brazed joint you never touched: that is equipment, and it routes to the manufacturer's warranty.
Route the manufacturer issues to the manufacturer instead of swallowing them. When the failure is a defect in the part, there is a claim to file to recover the part cost and often a labor allowance toward the swap, and the mechanics of doing that well live in the warranty claim processing guide. Tag every callback as workmanship or equipment when you log it. That single tag drives both the root-cause work, which is about your workmanship, and the recovery, which is about their parts.
Recovering the manufacturer's defects
When the part was defective, the manufacturer should pay for it, and a surprising amount of that money goes uncollected because nobody files. Registering the equipment at install, capturing the model and serial, and filing the claim when a part fails under its term is how you turn an equipment failure into a credit instead of a cost. Skip it and you donate the part and the labor to a defect that was never your fault.
Treat manufacturer recovery as a routine part of warranty cost management, not a chore. Every equipment-tagged callback is a candidate for a claim. The defective part comes off your shelf to keep the customer running, and the claim is how you get paid back for it, often with a labor allowance on top. Across a year, that recovery is real money that otherwise leaks straight out.
The full mechanics of registration, filing with proof, returning the defective part on a return authorization, fighting a denial, and tracking the credit until it posts are their own discipline, and they live in the warranty claim processing guide. Terms vary widely by manufacturer, so confirm the coverage, the term, and the labor allowance against the specific manufacturer's warranty language before you promise a customer anything. Here the point is narrower: separate equipment from workmanship, and recover the equipment side instead of absorbing it.
Honoring the obligation efficiently
When a callback is yours, honor it fast and honor it well, because a warranty visit handled right is one of the best sales calls you will ever make. The customer is watching how you behave when something goes wrong, which is when most contractors disappear. Show up quickly, fix it correctly, and be straight about what happened, and you have earned more trust than any marketing buys.
Efficiency matters on both sides of the ledger. Cluster warranty calls into routes so a tech is not burning a half-day of drive time for one fix. Send the right person with the right part the first time so the same callback does not turn into two visits. Fix the actual cause on site rather than patching it, because a warranty call you have to repeat is the most expensive kind there is. Done right, the visit costs you once and buys you the next job.
The relationship is the upside hiding inside the cost. A customer whose problem you solved cleanly under warranty is the customer who signs the maintenance agreement, calls you for the next system, and refers the neighbor. That conversion is real value, and the service agreement section comes back to it. The warranty visit is a cost you cannot avoid, so get the return on it.
Warranty versus a new paid job
The line you have to hold is between a real warranty obligation and a new, separate problem the customer wants fixed for free under the banner of warranty. Honor what you owe. Do not give away work you do not owe. Blurring that line is how a managed warranty cost turns back into a leak, because every freebie is margin you hand over with no record and no charge.
The pattern is familiar. The tech is on a legitimate warranty call, and the customer points at something unrelated: a different unit, an add-on they want, a part that failed from age or neglect rather than your work. It is easy to just take care of it while you are there, and that instinct is exactly what erodes the job. A failure from normal wear, from no maintenance, or from owner abuse is not your warranty. It is a new job to quote.
When the work is new, it gets a change order or a new proposal, priced and approved before you do it, the same as any other paid work. The change order guide covers that discipline in full. The tech needs the judgment and the backing to say, clearly and without apology, that the warranty fix is covered and the other item is a separate quote. Give them that, or you will keep paying for work you never agreed to do.
What the warranty does not cover
Every warranty has exclusions, and knowing them is what lets you hold the line in the section above without guessing. The usual ones are owner abuse, lack of maintenance, damage from other trades or other equipment, acts of nature, and normal wear on consumable parts. The exact list is in your warranty language, and it is worth reading before you are standing in a customer's mechanical room arguing about it.
The classic excluded callback is the no-maintenance failure. A filter never changed strangles airflow and overheats a system, a clogged drain backs up because nobody flushed it, a coil cakes over because it was never cleaned. That is a maintenance failure, not a defect in your work, and most warranties say so. Treating it as warranty trains the customer to expect free service forever.
The move is to document and communicate, not to argue on the spot. Note the condition you found, photograph it, and show the customer what actually failed and why it falls outside the warranty. Then offer the fix as paid work or, better, as a reason to put them on maintenance. Confirm what your specific warranty excludes against the terms, because the exclusions vary and a customer dispute is the wrong time to discover yours are narrower than you assumed.
How do you reduce warranty callbacks?
You reduce warranty callbacks by preventing them, and the cheapest callback is the one that never happens. Quality control, commissioning, and a real punch-list closeout before you leave the job catch the failures that would otherwise come back as warranty work weeks later. Money spent getting it right the first time is a fraction of what the callback costs, and it does not cost you a customer's confidence.
Commissioning is where most of this trade's callbacks get prevented or created. A system that was actually checked, charged to spec, balanced, and verified at startup does not generate the slow drip of nuisance calls that an install nobody finished does. The callbacks that look like equipment problems are often a startup that was rushed. The agent or the lead who confirms the system runs right before the truck leaves is buying down next quarter's warranty cost.
Close the loop back to the reserve and the root-cause data. The patterns you find in the callbacks tell you exactly which QC steps are worth adding, because they point at the details that fail. A shop that feeds its callback causes back into its install and commissioning standards watches its warranty percentage fall over time, which is the whole game. Prevention is not separate from cost management. It is the part that makes the cost go down instead of just getting covered.
Owner-caused failures and the false callback
A large share of callbacks are not your work failing. They are the owner's lack of maintenance failing, and they show up as warranty calls because the customer does not know the difference. A system that never gets a filter change, a drain flush, or a seasonal check will fail in ways that feel like a defect to the owner and look like neglect to anyone who opens the panel.
These false callbacks are reducible two ways. The first is education at handoff: tell the customer plainly what maintenance the system needs, what happens if it does not get it, and what is covered versus what is on them. A customer who understands the deal calls less and argues less. The second is a maintenance agreement, which puts you on the system on a schedule so the small problems get caught before they become a call.
The maintenance contract does double duty. It cuts the no-maintenance failures that drive false callbacks, and it converts the relationship into recurring revenue, which the next section covers. A customer on a plan is a customer whose system you keep running and whose callbacks you have largely designed out. That is cheaper than honoring the failures that neglect was always going to cause.
The reserve true-up
A reserve is an estimate, so it has to be checked against what actually happened and adjusted. Periodically you compare the warranty cost you reserved to the warranty cost you really spent, and you true it up: if you over-reserved, some can be released, and if you under-reserved, you add to it. A reserve nobody reviews drifts away from reality and stops doing its job.
The true-up is also your scorecard. If your actual warranty cost is falling because the root-cause work is paying off, the review is where you see it and where you can justify lowering the percentage you carry going forward. If it is rising, the review catches it before it becomes a year-end surprise, and it tells you to go find the cause. The number is a feedback loop, not a set-and-forget line.
How and when you true up a reserve, what you can release, and how any of it hits your statements are accounting decisions, and they sit with your CPA. Bring them the comparison of reserved versus actual on a regular cadence and let them handle the entries and the timing. Your job is to keep the data clean so the comparison means something. Theirs is to turn it into correct books.
Turning warranty relationships into service agreements
The upside buried in all this cost is that every warranty customer is a service agreement waiting to be signed. You already have the relationship, you already know the equipment, and you have already shown the customer how you behave when something breaks. Converting that into a maintenance agreement turns a pure cost relationship into recurring revenue, and it cuts the callbacks at the same time.
The pitch writes itself off the warranty visit. The customer just watched you fix their system cleanly under warranty. That is the moment to explain that a maintenance plan keeps the system running right, catches the small problems before they become emergencies, and protects the investment they made. You are not selling cold. You are extending a relationship that just proved its value.
The economics line up in your favor. The maintenance agreement reduces the no-maintenance failures that would have come back as warranty or service calls, so the customer is cheaper to keep happy. It adds predictable revenue between installs. And it keeps you on the equipment, so when the system finally needs replacing, you are the contractor who gets the call. The warranty obligation you have to carry anyway becomes the front door to the most profitable customer you have.
The records that make all of this work
None of this works on memory. The whole system runs on a callback log that captures, for every callback, the originating job, the cost, the cause, the crew, and the workmanship-or-equipment tag, plus a running comparison of reserved versus actual warranty cost. That record is what feeds the reserve, the pattern analysis, the root-cause fixes, the manufacturer recovery, and the true-up. Lose the record and you are back to flying blind.
This is where a field tool does the heavy lifting, because the record has to be captured at the point of work or it does not get captured at all. In FieldOS the callback is logged against the original job from the field, with its cost, cause, crew, and tag, so the callback rate and the warranty cost accumulate as real, sortable history. The office is not reconstructing it from a tech's memory a week later. It is already there, ready to size a reserve and rank by cause.
Keep the financial side of the record with the CPA. The log gives you the operational truth: what failed, what it cost, and why. The CPA turns the reserved-versus-actual numbers into the accrual and the statements. Both halves matter, and they feed each other, but do not confuse the field log with the books.
What to document
Capture the same facts on every callback so the history is sortable and the reserve has something real to stand on. Confirm the accounting treatment of anything reserve-related with your CPA, and confirm coverage questions against your warranty terms before you commit to a customer.
| Item | Action | Note |
|---|---|---|
| Originating job | Link the callback to the job that caused it | Without it, the cost cannot be matched back |
| Callback cost | Log parts, labor, drive time, and office time | The part alone understates the true cost |
| Cause | Record the actual root cause, not just the fix | This is what feeds the pattern analysis |
| Crew | Tag the crew that did the original work | Surfaces a training problem if one exists |
| Workmanship vs equipment | Tag which warranty applies | Routes equipment defects to a manufacturer claim |
| Manufacturer claim | File and track when the part was defective | Confirm coverage against the manufacturer terms |
| Reserve vs actual | Compare reserved to spent on a cadence | Confirm the treatment and true-up with your CPA |
| Exclusions found | Note and photograph owner-caused failures | Supports charging it as new work, not warranty |
Common mistakes
- Carrying no warranty reserve, so callbacks surprise the P&L in months that have nothing to do with the job that caused them.
- Not tracking what callbacks cost or why they happen, which leaves you with no history to size a reserve or find a pattern.
- Ignoring repeat callbacks and fixing the same failure again instead of escalating the process, training, or product cause behind it.
- Giving away non-warranty work as warranty because it is easier than telling the customer it is a separate quote.
- Never finding the root cause, so the same callback keeps coming back and you keep paying for it on every job.
- Not recovering manufacturer-defect costs, eating parts and labor allowances the manufacturer warranty would have paid back.
- Sizing the reserve off a number from an article instead of your own warranty history and your CPA's read of your books.
- Treating a no-maintenance or owner-abuse failure as warranty, which trains the customer to expect free service forever.
Field checklist
Want this checklist to run itself on every job — with photo proof and a signed record crews can hand the customer? That's FieldOS.
Standards, references, and where the professionals take over
The accounting side rests on revenue recognition and guarantee rules, and it is the CPA's ground, not yours. The standard known as ASC 606 governs how revenue and the warranties attached to it are recognized, and it draws a line between an assurance-type warranty that just promises the work was done right and a service-type warranty that sells the customer extra coverage. Assurance-type obligations are generally accrued under the guidance for guarantees. Which type yours is, and how the reserve is recorded and trued up, are determinations for your construction CPA against your actual books. Treat every percentage in this guide as illustrative.
The obligation side rests on your contract and your warranty terms. The common pattern of a roughly one-year workmanship warranty plus a longer manufacturer equipment warranty is a pattern, not a rule, and the difference between a callback period and a longer-running warranty of quality is a legal question that depends on your contract and your state's statutes. Read your documents, and take the hard questions to a construction attorney. Manufacturer coverage, terms, and labor allowances vary by manufacturer, so confirm them against the specific warranty language before relying on them.
On the quality side, the reduction work follows ordinary root-cause and Pareto practice: track the failures, rank the causes by cost, fix the top few at the source, and feed the lessons back into your install and commissioning standards. Three things carry the weight across all of it. Size the reserve from your own history, not a borrowed number. Track every callback and kill the root cause. And separate warranty from paid work while recovering the manufacturer defects you are owed. The change order guide covers holding the paid-work line, and the warranty claim processing guide covers the recovery mechanics.
Terms and definitions
Warranty cost management uses a handful of terms that get mixed up in conversation, so here is how this guide uses them. None of these definitions is accounting or legal advice, and the treatment of the financial terms is your CPA's call.
- Warranty reserve
- Money set aside against the warranty cost a job is expected to generate later, so the cost is matched to the job instead of surprising a future month. Also called a warranty accrual. Treatment is the CPA's.
- Callback
- A return visit to a job already closed, to fix a problem under warranty. Costs parts, labor, drive and office time, and the billable work the tech could have been doing instead.
- Accrual
- Recognizing an expected cost in the period the related revenue is earned rather than when the cash goes out, which is the matching idea behind a warranty reserve. How it is booked is the CPA's call.
- Root cause
- The actual underlying reason a callback happened, usually a crew, a detail, or a product, as opposed to the symptom you fixed on the visit. Fixing it removes future callbacks instead of just covering them.
- Workmanship warranty
- Your promise that the install was done correctly. Covers your labor and your details, commonly for about a year, but your contract controls the scope and the term.
- Equipment warranty
- The manufacturer's promise that the parts are free of defects for their term. Defective parts route to a manufacturer claim, often with a labor allowance. Terms vary by manufacturer.
- Service agreement
- A recurring maintenance contract that keeps you on the equipment on a schedule, cutting no-maintenance failures and false callbacks while adding predictable revenue between installs.
FAQ
What is a warranty reserve?
A warranty reserve is money set aside against the warranty cost a job will generate later, so the cost is matched to the job instead of surprising a future month. It is sized from your own history and trued up against actual cost. How it is booked is an accounting decision for your CPA.
What does a callback actually cost?
A callback costs far more than the part. Add the tech's labor, the round-trip drive time, fuel and truck wear, the office time to schedule it, and the biggest line of all: the billable work that tech did not do while fixing a job you already closed. A callback is never free.
How do you reduce warranty callbacks?
Track every callback by cause and cost, rank the causes, and fix the top few at the source through training, a process change, or a product switch. Prevent the rest with real commissioning and a punch-list closeout before you leave. The cheapest callback is the one that never happens.
What is the difference between workmanship and equipment warranty?
Workmanship warranty covers your labor and how you installed it. Equipment warranty covers the manufacturer's parts and whether they were defective. Workmanship failures draw from your reserve. Equipment defects route to a manufacturer claim for recovery. Tag every callback as one or the other when you log it.
How do you size a warranty reserve as a percentage of revenue?
Size it from your own warranty and callback history, broken down by job type, expressed as a percentage of the revenue that produced it. There is no universal correct percentage. The right figure comes from your numbers and your CPA's read of your books, and it moves as your quality changes.
How long is a contractor's warranty obligation?
A common pattern is roughly a one-year workmanship warranty plus a longer manufacturer equipment warranty, but that is a pattern, not a rule. A callback period and a longer-running warranty of quality are different things. Your contract, your warranty terms, and your state's law control, so read them and ask an attorney.
What does a warranty not cover?
Typical exclusions are owner abuse, lack of maintenance, damage from other trades, acts of nature, and normal wear on consumable parts. The no-maintenance failure is the classic example. The exact list is in your warranty language, so confirm it against your terms before treating a callback as covered or excluded.
How do you stop giving away non-warranty work as warranty?
Hold the line between a real warranty obligation and a new problem the customer wants fixed for free. A failure from age, neglect, or abuse is not your warranty. Document it, show the customer, and quote it as new work with a change order before doing it, the same as any paid job.
What should you track on every callback?
Log the originating job, the full cost, the root cause, the crew, and a workmanship-or-equipment tag, plus a running comparison of reserved versus actual warranty cost. Captured in the field at the point of work, that history feeds the reserve, the pattern analysis, the root-cause fixes, and the manufacturer recovery.