Revenue tells you the shop is busy. Five numbers tell you whether it survives: gross margin per ticket, average ticket, labor utilization, callback rate, and months of cash reserve. Most owners track none of them consistently. Each can be captured in under an hour a week, and together they show exactly where the shop is leaking money. Part [N] of the nine-part Fix the Shop First series.

Fix the Shop First · Part [N] of 9

By Rob Link, Founder & CEO, Tech Care Association · September 16, 2026 · 20 min read

Key takeaways

• Busy and profitable are different things. Shops close with a full bench, and the closure record shows it.
• Five numbers cover the whole picture: gross margin per ticket (finance), average ticket (pricing), labor utilization (capacity), callback rate (quality), and months of cash reserve (survival).
• Each one has a one-line formula, and every input already exists in your point of sale, your payroll, or your bank statement.
• Capture all five for last month as a baseline. Pick one to improve. Re-measure next month.
• If you take insurance claims or third-party warranty work, run the five numbers for that channel on its own. Program jobs produce gross revenue; they do not always produce margin.
• TCA ShopCheck gives you the same baseline in about two minutes. This chapter gives you the detail behind it.

What shops are saying

Observational evidence from TCA Repair Community Listening. Method and limits are in the footer.

In the online discussions the TCA reviewed this week, repair professionals argued about price in nearly every thread: what to charge for labor, what to charge for a screen, what to pay a technician. Almost none discussed what the shop keeps after the repair. Across the discussions reviewed, one participant stated a margin figure. No participant mentioned reserves, runway, or months of expenses in the bank.

That is not a lack of interest. Price is the number on the receipt. Margin is the number nobody handed them.

The most contested topic was economy LCD screens versus OLED, which surfaced in four separate discussions in one week. One camp argued that installing economy screens damages a shop's reputation for a small one-time profit, and that the customer who "never comes back" went somewhere else. Another camp, including an owner who once sold OLED only, argued that customers choose the economy option when it is offered, that people who break phones break the next screen regardless, and that shops should stop running each other down. A third position, held by several, was that the choice is fine as long as the customer understands what they are buying.

Every side argued from anecdote. Nobody cited a warranty-return number. Hold that thought for number four.

Why "busy" is the most dangerous word in repair

Busy feels like safety. The phone rings, the bench is full, the deposit goes in every day. Then a shop that was full every day closes, and everyone who knew the owner says the same thing: they seemed to be doing fine.

The TCA has written about this twice. Why Tech Repair Shops Fail laid out the closure record. Bureau of Labor Statistics data, as analyzed by Commerce Institute in 2025, puts first-year failure for retail businesses at 15.8 percent and five-year failure at 41.7 percent. The TCA's own estimate for independent tech repair is worse: first-year failure close to 30 percent, nearly double the retail figure, and an annual failure-or-exit rate of 30 to 40 percent. That article also listed what survivors do differently, and one habit on the list was tracking their numbers. This chapter is the how.

Nobody knows the exact number of independent tech repair shops in the U.S. The TCA works from roughly 40,000, give or take. At a 30 percent annual exit rate, that is more than 10,000 shops closing every year, and as Why Tech Repair Shops Fail put it, "The industry is running a 100% replacement rate. Every year. Just to stay flat."

The Data Gap explained the cost of running a shop on numbers you do not have. The TCA's estimate there was that the average independent shop leaves $50,000 to $75,000 on the table every year through pricing errors, dead inventory, missed repeat business, and supplier problems it never measured. You cannot fix pricing you never measured.

The wider small business data says the same thing in a different accent. Across all U.S. industries, Bureau of Labor Statistics data puts first-year business failure at 20.4 percent and five-year failure at 49.4 percent. In a February 2026 Bluevine survey of 785 small business owners, 56 percent said they hit significant cash flow problems within their first three years, 26 percent said financial tasks take more time than they expected, and bookkeeping was the biggest weekly time drain for 42 percent. And when CB Insights analyzed 431 venture-backed startups that shut down since 2023, 70 percent had run out of capital, but the report's own conclusion was that running out of money is where the story ends, not why it ends. Unsustainable unit economics, meaning the business cost more per customer than it earned, showed up in 19 percent of those post-mortems. Those were funded startups, not repair shops. The lesson transfers anyway: the bank balance is the symptom. The five numbers are the diagnosis.

The community discussions this week put faces on all of it.

One participant described holding the same labor price for two years, considering an increase, and summed up his shop as not high-volume but consistently busy. Busy but flat, in one sentence.

Another participant, in a discussion about whether $80 is too little to charge for a screen replacement, put his overhead beside that figure. With combined housing and shop lease costs above $6,000 a month, he observed, $80 labor only works if the shop turns a repair every ten minutes. That is the whole chapter in one comment. The price is not wrong on its own. The price is wrong for that shop, at that overhead, at that throughput. You cannot know which one you are until you do the math.

A third participant, replying to a thread full of stated labor prices, asked in effect whether anyone was getting paid at all. It was a joke. It got a lot of agreement.

And two participants said the chain store in the next town caps what they can charge. One said he cannot go above $85. When the ceiling is set from outside, the only thing left to control is the floor, and the floor is set from inside. That is what the five numbers are for.

Number 1: What am I actually keeping on each repair?

This is the number that separates a busy shop from a profitable one. Revenue is what the customer paid. Gross margin is what was left after you paid for the part and paid the person who installed it, before rent, software, insurance, and everything else that keeps the lights on.

Where each input lives:

• Ticket revenue: your point of sale. Total repair revenue for the month, excluding retail and accessory sales if you can separate them. If you cannot, note it and be consistent every month.
• Parts cost: your supplier invoices or your POS if it tracks cost of goods. Use what you actually paid, including shipping.
• Direct labor cost: payroll for technician hours only, at their loaded rate (wages plus employer taxes). If you are the technician, pay yourself a rate on paper for this line, or the number lies to you.
• Tickets: your POS count of completed repair tickets.

A worked example (illustrative numbers, not a benchmark): a shop closes 120 repair tickets in a month for $14,400 in repair revenue, spends $5,200 on parts and $3,600 on direct technician labor. Gross margin is $5,600, or $46.67 per ticket. That $46.67 has to cover every other cost in the shop before there is a dollar of profit.

Subtract what your own warranty costs. Gross margin per ticket is what a repair leaves behind on the day. Some of that money is already spoken for, because a share of tickets will come back under warranty and you will do the second repair for free. The expected cost is simple: callback rate (number four) multiplied by what an average callback costs you in part, labor, and counter time. On the illustrative example, a 3.3 percent callback rate against a $90 average callback is about $3 per ticket, which is manageable. On a $30 economy install with a higher return rate, the same math can eat the whole ticket. Write the number down; it is the price of the promise on your receipt.

Run it by channel. If any of your tickets are paid by someone other than the customer in front of you, split the number. Walk-in repairs, insurance claims, third-party warranty program jobs, and business contracts each carry different revenue, parts rules, labor, and admin time, and a blended average hides which one is carrying the others. The section after number five walks through why.

Setting a target: the TCA is not going to hand you a margin target from a source that does not exist. Set your own: take last month's gross margin per ticket, then multiply by your monthly ticket count. If that total does not cover last month's fixed costs with room to spare, your target is the per-ticket number that would.

Number 2: What is my average ticket, and is it moving?

This is the easiest of the five to pull and the one most owners already half-know. Track it monthly and watch the trend, not the figure. Flat average ticket with rising parts cost is falling margin, even when revenue looks fine.

Where it lives: your POS. Same revenue and ticket count as number one.

Both are levers. Which one is right for your shop depends on your margin per ticket (number one) and your callback rate (number four). The December case study in this Guide is built around exactly this number, so record it now.

Number 3: How much of the day is billable?

You pay for every hour a technician is in the building. You only get paid for the hours that land on a ticket. The gap between them is diagnostics that did not convert, waiting on parts, counter time, cleanup, and the customer who wanted to talk for twenty minutes.

Where it lives: paid hours from payroll or your schedule; billable hours from your POS if it records time per ticket, or from a simple estimate per repair type multiplied by tickets completed. A rough number tracked every month beats a precise number tracked never.

Why 100 percent is impossible: nobody bills every paid minute. A technician who is at the bench all day still has intake, reassembly checks, part lookups, and the two-minute questions that turn into ten. Utilization is not a grade. It is a ratio you watch so you can tell whether a slow month was slow because the phone stopped ringing or because the bench got inefficient.

Setting a target: the TCA does not have a sourced industry range for repair-shop utilization and will not invent one. Measure yours for three months. The best of the three is your first target.

A worked example (illustrative): a technician is paid for 160 hours in a month. Tickets completed, at the shop's own time estimates, add up to 88 billable hours. Utilization is 55 percent. At that shop's labor rate, the owner can now see what the other 72 hours cost.

If you are the only technician: several participants in the pay discussion were solo operators. One wrote that he does not pay himself. Another asked whether people really have employees. Utilization still applies. The paid hours are your hours in the shop, and the number tells you how much of your week is actually earning.

Number 4: How often does a repair come back?

A callback is not one cost. It is the second part, the second block of bench time, the technician pulled off a paying ticket, and a customer who now tells the story at dinner. On a thin-margin ticket it removes the whole margin. On a premium ticket it removes a slice of the margin and adds a slice of doubt.

Where it lives: your POS, if you flag warranty work as its own ticket type. If you do not, start Monday. Count every device that returns for any reason related to a prior repair, whether or not you decide it was a defect.

Count the disputes too. In a discussion about customer excuses for damage, owners cataloged post-repair blame: a screen that cracked on a short drop onto carpet attributed to a cheap part, battery and charging-port trouble attributed to a screen repair, a phone that supposedly fell off a nightstand arriving with damage consistent with being run over. Those are not defects. They still cost bench time, and the callback rate should capture them. Split the number into "our fault" and "not our fault" later if you want; get the total first.

The TCA is not going to settle that argument, and neither are you, until warranty returns divided by completed repairs is a number your shop tracks by part tier. At that point it stops being an argument.

Number 5: How long could the shop run on what is in the bank?

This is the survival number. It answers one question: if revenue stopped tomorrow, how many months would the shop stay open? A slow season, a lost lead technician, a landlord who does not renew, a supplier who changes terms. Each one is survivable with a reserve and fatal without one.

Where it lives: cash on hand from your business bank balance (not counting money already owed to suppliers or payroll due this week). Money an insurer or warranty program owes you is not cash on hand until it lands; track it separately as receivables, with the days it takes to arrive. Fixed costs from the last three months of bank statements: rent, utilities, software, insurance, loan payments, base payroll. Average them.

Setting a target: there is no sourced industry figure for how many months a repair shop should hold, and the TCA will not invent one. In Why Tech Repair Shops Fail, the TCA's self-assessment asked whether a shop has three months of operating expenses in reserve. Treat that as the TCA's rule of thumb, not a benchmark, and treat your calculated baseline as the number that matters. If the answer is under one month, that is the number to fix before any other project in this Guide. Set a target you can name, and put a fixed amount toward it every month before you spend on a new tool or a new location.

A worked example (illustrative): $18,000 in the business account against average fixed costs of $9,000 a month is two months of reserve.

Before you say yes to insurance and warranty program work

Somebody offers you volume. A device protection program, an insurance carrier's repair network, an extended warranty administrator, a mail-in fulfillment partner, a corporate fleet. Claims get routed to your shop, the bench stays full, and monthly revenue jumps. It looks like the thing every owner wants.

It can be. It is also the fastest way the TCA knows to build a busy shop with no margin, because program work quietly removes several of the levers you just read about.

Where the net goes:

• You no longer set the price. Reimbursement is fixed by the program's fee schedule. Average ticket (number two) stops being a lever you control, and the add-on question that moves it does not exist on a claim.
• You may not choose the part. Many programs specify the part grade or the supplier. Your parts cost is what they say it is, not what you negotiated.
• Admin time does not bill. Claim portals, intake photos, pre-approvals, status updates, disputes, and resubmissions are real hours at the counter that never land on a ticket. Utilization (number three) drops and nobody sees it, because the bench still looks busy.
• You get paid later. Program terms are often net 30, 45, or 60 days. You have already paid for the part and the technician. Until the remittance arrives you are financing the insurer, and that shows up in months of cash reserve (number five).
• Claims get denied and clawed back. A rejected claim after the repair is done is a full-cost ticket at zero revenue.
• You carry their warranty at your cost. Most programs require the shop to honor a warranty on the repair, often longer than the one you offer walk-ins, and the callback is usually on you. Your callback rate (number four) on program jobs is a direct cost with no upside.
• The customer is not yours. They belong to the program. The repeat visit, the review, the referral, and the accessory sale rarely come back to your shop.

The test is the same five numbers, run for that channel alone. Take a month of program tickets. Reimbursement received, minus program-specified parts, minus direct labor, minus the admin hours at your technician's loaded rate, minus the expected warranty cost (the program's callback rate times what a callback costs you). Divide by program tickets. That is your true margin per program ticket. Multiply by monthly program volume and put it beside the walk-in margin those same bench hours could have earned. Then ask how many weeks of payables you are covering while you wait to be paid, and whether your cash reserve absorbs it.

A worked example (illustrative numbers, not a benchmark): a program pays $180 for a screen claim. The required part costs $95, direct labor is $30, and the technician spends 20 minutes on the claim portal at a loaded $25 an hour, call it $8. That leaves $47 before warranty. If program jobs come back at 5 percent and a callback costs the shop $110 in part and labor, expected warranty cost is about $5.50, so the ticket nets roughly $41. Now hold that $41 against a payment that arrives in 45 days and a walk-in ticket that pays today. Whether that works is not a matter of opinion. It is a matter of which number is bigger for your shop.

Do they make enough to support the warranty? Only the channel math answers that. If the true margin per program ticket is positive after admin time and expected callbacks, and the payment lag fits inside your cash reserve without squeezing suppliers or payroll, the work stands on its own and can be a sensible way to fill idle bench hours. If the margin only looks positive before admin and warranty are counted, or if the receivable lag is being carried by your walk-in cash, you are not running a repair program. You are lending the insurer money at zero percent and paying for the privilege with your bench.

The TCA's position: program work is a filler, not a foundation. Take it when it clears its own margin and you have the reserve to wait for payment. Never build the shop's revenue base on it, because the day the program changes its fee schedule or its network, that revenue leaves the same way it arrived.

How to capture all five in under an hour a week

You do not need new software. You need one sheet and one recurring hour. The Five Numbers worksheet is a free PDF with the nine inputs, the five formulas, and a target line for each; print it or fill it in on screen.

Weekly (about 30 minutes):

1. Pull from your POS: repair revenue, tickets completed, warranty tickets. Three numbers.
2. Pull from your parts invoices: parts spent this week. One number.
3. Pull from your schedule or timeclock: paid technician hours. Estimate billable hours from tickets completed if your POS does not track time. Two numbers.

Monthly (about 20 minutes, on top of the last weekly session):

4. Pull from the bank: cash on hand and the fixed-cost total for the month. Two numbers.
5. Pull from payroll: direct technician labor cost for the month. One number.

Nine inputs, five formulas, five answers. Keep every month's sheet and the trend builds itself. If you take insurance or warranty program work, the second page of the worksheet runs the same numbers by channel, with a line for receivables and expected warranty cost. The first month is your baseline. The number you fill in first should be the one you found yourself arguing about this week.

What to do Monday morning

1. Open the Five Numbers worksheet and fill in last month. Nine inputs, five outputs. Do not chase precision on the first pass; chase completion.
2. Circle the one number you were arguing about with someone this week. Margin, price, hours, callbacks, or cash. That is the one you improve first. Pick one lever from the matching section above and write down what you will change.
3. Put the same hour on next Monday's calendar. Re-measure. The second month is where the trend starts and the guessing stops.

Busy is a feeling. Profitable is a number.

Every owner in the discussions the TCA reviewed this week could tell you their labor price. Almost none could tell you what a ticket leaves behind, how much of the bench is billable, how often a repair comes back, or how many months the shop could run on what is in the bank. Those are not accounting questions. They are the difference between a shop that is busy and a shop that survives, and every one of them takes less than an hour a week to know.

Next Tuesday, the series turns to pricing labor. Do not start that chapter with a rate. Start it with these five numbers filled in for last month.

Common questions about repair shop profitability

What is gross margin per ticket for a repair shop?

Gross margin per ticket is the money left on an average repair after the part and the direct technician labor are paid, before rent and other overhead. The formula is (repair revenue minus parts cost minus direct labor cost) divided by tickets completed. It is the number that separates a busy shop from a profitable one.

What is a good callback rate for a tech repair shop?

The TCA has not published a benchmark for callback rate and does not recommend using one from an unsourced list. Calculate yours as warranty returns divided by completed repairs, including disputes that are not defects, then track it by part tier. Your first target is your own best month.

How many months of cash reserve should a repair shop have?

There is no sourced industry figure. The TCA's rule of thumb, from Why Tech Repair Shops Fail, is three months of operating expenses. Calculate your own as cash on hand divided by average monthly fixed costs. If it is under one month, that is the number to fix before any other project.

Should a repair shop take insurance or warranty program work?

Only when it clears its own margin. Run the five numbers for the program channel alone: reimbursement minus required parts, direct labor, unbilled admin time, and expected warranty cost, divided by program tickets, then check whether the payment lag fits inside your cash reserve. Program work is a filler for idle bench hours, not a foundation for the business.

How long does it take to track these five numbers?

Under an hour a week: about 30 minutes weekly for six inputs from your POS, parts invoices, and schedule, plus about 20 minutes monthly for three inputs from the bank and payroll. The free Five Numbers worksheet has the formulas printed beside each input.

More from the TCA Blog

• Fix the Shop First: Stop Chasing the Next Thing
• Why Tech Repair Shops Fail
• The Data Gap Costing Tech Repair Shops $50,000+ Per Year
• How Healthy Is Your Repair Shop? Take the Free 2-Minute Health Check
• Data Recovery: Should Your Shop Offer It?

Rob Link is the founder of the TCA. He built and ran a multi-location repair chain in an earlier era of this industry, and he'd tell you today's operators have it harder. The TCA is independent of carriers, manufacturers, and insurance companies, which is exactly why it can tell shop owners the truth about growth.

Sources and methodology

Published September 16, 2026. Last updated September 16, 2026.