Every automotive process in your shop was designed by somebody. Usually that somebody was a chaotic Tuesday in 2014, two techs quitting the same week, and a service advisor who said, “just do it like this for now.” Nobody wrote it down. And “for now” is still running your parts counter.
Here’s what that costs. Across a busy shop, undocumented processes leak money in small, boring increments — an untraceable part here, a core that never got credited there, a vendor overbill nobody caught. It rarely shows up as one big loss. It shows up as a margin number that’s a couple points softer than it should be, every single month, and nobody can quite say why.
This is a playbook for the owner or MSO who runs the operation, not the consumer paying the bill. It covers the seven processes worth writing down, the ones where the money actually moves, and how to build an SOP your crew will follow after you’ve left for the day. Read it in five minutes. Roll out the first SOP this weekend.
Why “the way we’ve always done it” is the most expensive process in your shop
“We’ve always done it this way” is one of the most expensive phrases in auto repair. It sounds like stability. It’s usually just an undocumented process nobody has questioned since a Bush was in office.
Here’s the problem with “always done it this way.” It isn’t one way. It’s four ways, because four different people each learned it from someone slightly different, and none of it is written down. The invoice gets filed here on Mondays and there on Fridays. Cores go in the blue bin unless the blue bin is full, in which case, who knows. Your process didn’t get designed. It accumulated.
An undocumented process is a process that walks out the door when your best parts guy takes a job across town. Everything he “just knew” leaves with him. The new person invents their own version, which is subtly worse in ways you won’t notice for six months. That’s not a people problem. It’s a process problem, and it’s fixable.
Process beats memory. Every time. If your shop’s operations live in three people’s heads, you don’t have a system — you have three single points of failure who occasionally take vacations.
The 7 core processes every repair shop should have written down
You don’t need a binder with 90 SOPs. You need the handful that touch money and repeat every day. Here are the seven that earn their spot.
- Parts ordering and PO numbers — how parts get ordered and how each purchase ties back to a job.
- Returns, cores, and credits — how a returned part becomes actual money back in your account.
- Estimate → authorization → invoice — the paper trail from quote to signed approval to final bill.
- Daily cash and deposit balance — what the drawer and the shop management system should agree on before the deposit leaves.
- Month-end AP and reconciliation — how you confirm every dollar spent on parts before you pay a vendor.
- Warranty and comeback handling — how a redo gets tracked, coded, and kept out of your gross profit.
- New-hire onboarding — how a new person learns all of the above without inheriting somebody’s bad habits.
The four that leak the most cash — parts ordering, returns and credits, estimate-to-invoice, and month-end reconciliation — get their own sections below. Nail those first. The rest can wait for the following weekend.
The parts-ordering and PO-number process (where margin leaks start)
If you fix one automotive process this year, fix parts ordering. It’s the first domino. Everything downstream — reconciliation, credits, inventory, margin — depends on whether a part can be traced back to the job that needed it.
The rule is simple: the repair order number is the PO number. Every part ordered for a specific job goes on the vendor invoice with that RO number attached. No blanks. No “misc.” No made-up numbers that mean nothing by month-end.
Why does this one convention matter so much? Because it gives every purchased part an ending. A part you bought has to land somewhere — sold on the RO, put into inventory, or returned and credited. When the RO number rides on the invoice, you can trace any charge to a job in seconds. When it doesn’t, a bought-not-sold part just dissolves into COGS and nobody ever asks where it went.
I think about Bob, my co-founder, who owned shops long before we built software. About a decade ago he discovered more than $180,000 had walked out over time through a technician and a service advisor ordering duplicate parts and working the gaps in the process. It wasn’t one dramatic theft. It was a slow leak through a hole in the ordering process — no clean tie between a part bought and a job that needed it. The lesson was never “people steal.” It was that a tight ordering process protects every shop, including the honest ones.
For the specific habits that drain a parts counter, we broke down the most common parts-ordering mistakes costing your shop money. Write your ordering SOP so those mistakes are physically hard to make.
Illustrative dollar math: the untraceable-PO leak across a 5-shop group
Say you run five locations, each closing 300 repair orders a month. That’s 1,500 ROs across the group every month.
Now assume your PO process isn’t standardized — some invoices carry a clean RO number, plenty don’t. Conservatively, that gap lets just $7 per RO leak through: a bought-not-sold part that never got questioned, a small overbill nobody matched, a credit that was owed but never chased.
$7 × 1,500 ROs = $10,500 a month. That’s $126,000 a year — from one weak convention, spread so thin across so many tickets that no single store ever loses enough to notice. (Numbers are illustrative; plug in your own RO count and it still stings.) At a group level, small leaks don’t stay small. They just get quiet.
The return, core, and credit process (a return isn’t done until the credit lands)
Here’s the sentence to tape above the parts counter: a return isn’t done until the credit lands. A return slip is not money. An empty box on the truck is not money. A credit memo posted to your account is money.
Most shops treat a return as finished the moment the part leaves the building. Financially, it’s barely started. The part still has to be received by the vendor, accepted as returnable, and credited — and any one of those steps can fail without anyone noticing. Cores are worse, because a core deposit sits there looking like a normal charge until the return window closes and it becomes a donation.
The process has three checkpoints, and the SOP has to name all three:
- The part physically goes back — logged, on a return slip, with the driver’s signature on your copy.
- The credit is expected — recorded as owed, tied to the original RO and invoice.
- The credit is confirmed — matched against the vendor statement before you consider it closed.
Skip the third checkpoint and you get the classic leak: everyone remembers sending the part back, nobody confirms the money came back. We wrote a whole piece on how the return-and-forget loophole destroys your bottom line — it’s the return process failing at that last checkpoint, on repeat. The catch happens at vendor statement reconciliation, which is exactly why the credit process and the month-end process have to talk to each other.
Cores aren’t homing pigeons. They don’t come back and turn into credits on their own. Somebody has to own the ending.
The estimate → authorization → invoice workflow
This process protects two things at once: your margin and your right to get paid. It’s the paper trail that runs from “here’s what it’ll cost” to “you approved it” to “here’s the bill.”
The workflow has three non-negotiable states, in order:
- Estimate — a written, itemized quote with parts, labor hours, and your rate, before any wrench turns.
- Authorization — the customer’s approval on record, including a documented approval for any supplement that pushes the total past the original number.
- Invoice — the final bill that matches what was authorized, line for line.
The leak lives between authorization and invoice. A tech finds extra work, the advisor “mentions it” over the phone, the customer sort of agrees, and the labor goes on the RO without a documented yes. Now you’ve got a discount you didn’t decide to give, or a dispute you can’t win, because the authorization for that line doesn’t exist in writing.
Two numbers make or break this workflow: the labor rate on the estimate and the parts price on the invoice. Get the rate wrong and every authorized hour leaks margin — worth confirming you’re charging the right labor rate for your shop in 2026. The SOP’s job is to make sure nothing gets billed that wasn’t first estimated and authorized, and nothing gets authorized that isn’t priced right.
The month-end AP and reconciliation process
This is where all the other processes get graded. Every part you ordered, returned, and credited during the month lands here, on a vendor statement, waiting to be confirmed before you pay.
Here’s a hot take that saves shops real money: if your AP process only starts when the statement arrives, you’re already behind. By the time that statement shows up, you’re weeks late to every problem on it. A missing invoice becomes a fire drill instead of a five-minute fix. A missing credit becomes an argument you’ve probably lost.
The better version is invoice-first. You work invoices and credits throughout the month, resolve exceptions while the vendor’s dispute window is still open, and use the statement as the final check before payment — not as the first time anyone looks. The month-end SOP then reads like a checklist, not a panic:
- Weekly: match each parts invoice, line by line, against the RO and packing slip. Flag overbills, wrong quantities, and untraceable POs now.
- Weekly: confirm expected credits actually posted. Chase the ones that didn’t.
- Month-end: reconcile each vendor statement against your confirmed invoices and credits.
- Month-end: only pay a vendor once every line is supported.
This is the one process where software genuinely earns its keep, and it’s honest to say so: WickedFile runs your month-end reconciliation SOP for you, matching vendor invoices, credits, and statements against your repair orders and flagging the exceptions — the missing invoice, the unapplied credit, the part bought but never sold. What it does not do is write your other six SOPs, order your parts, or authorize a customer’s work. It enforces one process. The rest are still on you. And if a clean PO convention plus daily invoice uploads already gets your month-end where it needs to be, you may not need software at all — plenty of disciplined single shops don’t. For the bigger picture on where the money goes, we mapped out how to catch profit leaks before they drain you.
Standardizing processes across multiple locations (the MSO problem)
One shop with a weak process leaks a little. Five shops each running their own version of a weak process leak in five different directions, and you can’t compare any of it. That’s the MSO problem in one sentence.
When you add a location, the informal stuff breaks first. “The owner eyeballs the invoices” doesn’t scale past the store the owner is standing in. Each shop negotiated its own vendor terms. Each parts guy codes purchases his own way. By store number three, your consolidated P&L isn’t comparable, because you’re adding up numbers that were built five different ways.
The fix is to standardize the money-touching processes at the group level and let the rest stay local:
- One parts-ordering convention — same PO rule at every store.
- One chart of accounts and location code — so the roll-up actually adds up.
- One month-end close checklist — same steps, same order, every location.
- One named owner of reconciliation — a real person accountable across all stores, not “the back office” in the abstract.
Everything that doesn’t touch a dollar — how the lobby coffee gets made, which bay does alignments — can stay local. Standardize the money. Leave the personality alone. We go deep on the operation itself in the multi-location auto repair back office playbook.
How to write, roll out, and audit an SOP people actually follow
An SOP nobody follows is just expensive wallpaper. The goal isn’t a beautiful binder. It’s a written process your crew uses when you’re not watching. Here’s how to get one that sticks.
Write it (keep it to one page). The person who does the job writes the draft, in plain shop language. Name the trigger — when the process starts. List the steps in order. Say who owns each one. Explain the why, not just the what; people follow a rule they understand and quietly ignore one they don’t. AutoOps has a solid walkthrough of what a clean auto-repair SOP should contain, and Ratchet & Wrench’s long-running case for SOPs makes the same point from shops that have run them for decades.
Roll it out (test before you trust). Hand the draft to the newest person on the team and watch them run it cold. Where they get stuck is where your SOP is unclear — fix that, not them. Then make it the standard out loud: this is how we do it now, here’s why, questions welcome.
Audit it (or it rots). Every process drifts. Once a quarter, compare what the SOP says against what’s actually happening on the floor. If reality has moved, either update the SOP or fix the drift — but don’t let the two silently disagree, because a stale SOP teaches everyone that the SOPs are optional.
Trust your people. Verify your process. Those aren’t in tension — the second one is how you keep earning the first.
If your shop’s processes currently live in the heads of three people who all take vacation the same week, pick one SOP from this list and write it down before your margin needs its own filing cabinet.
