There is one number that separates appliance repair businesses that make money from ones that stay busy and broke, and it is not average ticket. It is the percentage of jobs finished on the first visit.
The math is unforgiving. A residential appliance call carries roughly an hour of billable labor and a parts markup. If you have to come back with a part, you spend the drive time twice, you set up twice, and you burn a slot on the calendar that another customer would have paid for. On most jobs the second trip eats the entire margin. Do that on a third of your calls and you are running a business that generates revenue and no profit, which is exactly how a shop can be booked out three weeks and still be short at the end of the month.
As of August 2026 the parts side of this trade is harder than it used to be, not easier: more electronic control boards, more brand-specific assemblies, and more units where the correct part hinges on a serial-number break mid-model-year. That makes the workflow around the metric — what you capture at booking, what history the tech can see, what actually rides in the van — the whole ballgame. This guide covers that workflow and how to measure the result without lying to yourself.
Why the second trip costs more than it looks
Owners tend to price the return visit as "just the drive." It is more than that.
There is the drive both directions, which in a spread-out service area is often longer than the repair itself. There is the second setup: pulling the unit out, dropping the back panel, re-confirming the diagnosis. There is the displaced job, which is the real cost — a slot that could have been a fresh call with a full diagnostic fee attached. There is the phone time to schedule the return and the follow-up when the customer is not home. And there is the risk that between visits the customer calls someone else who has the part on their truck.
Add it up and a return visit on a job with an hour of labor is usually a losing transaction, before you have even considered that you probably will not charge a second service call fee for a part you failed to bring. If you want to see it as a line rather than a feeling, run true profit per job with the actual labor hours and drive time on the callbacks, and the picture gets very clear very quickly.
Capture the model and serial at booking, not at the door
The single highest-leverage change most shops can make costs nothing: get the model and serial number when the appointment is made instead of when the technician is standing in the kitchen.
A brand and a symptom is not enough to prepare. "Whirlpool dryer, not heating" describes a dozen possible parts across several platforms. A model number narrows that to one parts breakdown, and a serial number resolves the mid-model changes that otherwise send you back for the correct variant of a part you already brought.
The practical version of this at intake:
- Ask for brand, model, serial, and approximate age on every call.
- If the customer cannot read the plate, text them and ask for a photo of it. Most people can find a sticker inside the door, behind the drawer, or on the back panel in about ninety seconds.
- Write the symptom in the customer's words, then in the technician's words. "Makes a grinding noise on spin" is more useful than "washer broken."
- Note the install location and access — second floor, stacked unit, tight laundry closet — because that changes what tools go on the truck.
That record has to be attached to the job, not to a sticky note at the front desk. The failure pattern here is not that nobody asks, it is that someone asks and the answer never reaches the technician. This is the same discipline behind good repair order retrieval: the information is only worth capturing if the person doing the work can pull it up in five seconds on the way to the job.
Once you have a model number before dispatch, a decent number of calls stop being diagnostic visits at all. If the symptom on that model points at a known common failure, the part rides along on the first trip and the visit closes.
Repair history per appliance, not per customer
Most software tracks customers. Appliance repair needs something narrower: a record attached to the specific unit.
A house may have five appliances. The washer that failed in March and the dishwasher that failed in October have nothing to do with each other, and lumping them into one customer note makes the useful signal disappear. What a technician actually wants to see standing in front of a unit is: has anyone been inside this machine before, what did they replace, when, and does it carry warranty.
That record changes behavior in three concrete ways. First, a repeat failure on the same unit is usually connected to the first repair, and knowing what was replaced fourteen months ago starts the diagnosis in a different, better place. Second, it prevents the awkward conversation where you charge a customer for a part your own shop installed last year and should be covering — an outcome that is bad for the relationship and worse for the review. Third, it makes manufacturer and extended-warranty work tractable, because you can retrieve what you installed and when without hunting through paper; that lookup is the foundation of appliance claims that actually get reimbursed instead of aging out.
Practically, this means the appliance is a record with a brand, model, serial, install date if known, and a service history under it, and the customer is the parent record holding one or more appliances. If your system cannot represent that, technicians will keep it in their heads, and heads do not transfer between employees.
Let the parts you fetch twice drive your stocking
Every supplier will sell you a recommended van kit. Ignore it and use your own data instead, because your mix of brands and your service area's housing stock are not the national average.
The report you want is simple: over the last six months, which parts were ordered after a diagnostic visit? That list is a direct measurement of the parts that cost you second trips. The top of it is your standing van kit. For most residential shops it clusters in a predictable place — door switches, drive belts, water inlet valves, lid switches, thermostats, igniters, drain pumps — but the brands and the exact variants will be specific to your market.
Then apply the two-population rule that governs all field-service stocking. Cheap, fast-moving parts get a reorder point set around lead-time consumption plus a buffer, because the failure mode is a stockout and stockouts are expensive. Expensive, slow-moving parts — control boards especially — get a reorder point of one or zero with an order raised when the job books, because the failure mode there is dead capital and obsolescence. A $340 board for a model you touch twice a year is not inventory, it is a bet.
Two more habits keep the kit honest. Track stock per van, not company-wide, so a technician can confirm the part is in the vehicle they are driving rather than the one parked at the shop. And watch what the kit does to margin: parts that sit for a year quietly erode the numbers behind your parts markup and margin, which is why "stock everything" is not a strategy. The deeper version of the stocking question is worked through in the appliance repair parts inventory guide.
Measure it honestly, or do not measure it
Here is where most shops quietly fail. They define first-visit fix rate as jobs completed on visit one divided by total jobs, and then they start carving out exceptions: the customer was not home, the part was on national backorder, the unit was not the model they told us. Every exclusion makes the number nicer and the metric useless.
The correct formula is the blunt one:
First-visit fix rate = jobs closed on the first visit ÷ total jobs dispatched.
Count the ordered-part jobs as misses. That is not an accusation against the technician; it is the point of the measurement. The rate exists to make the ordering visible, and if you exclude the ordering you have built a number that cannot move.
Cut it three ways once you have a few months of data:
- By technician. Two techs with the same territory and very different rates usually differ in diagnostic thoroughness or in how they stock their van, and both are coachable.
- By appliance type. Refrigeration and front-load washers usually trail dryers and ranges. Knowing by how much tells you where the parts investment goes.
- By source of the miss. Wrong part, no part, misdiagnosis, access problem, customer not home. This is the column that generates actions.
Pair it with callback rate — jobs that came back after you thought they were done — because a shop can juice first-visit numbers by declaring jobs complete that are not. The two metrics together are honest; either one alone can be gamed. The callback and rework rate breakdown covers that second half in detail.
| Cause of the second trip | What it costs | The record that prevents it |
|---|---|---|
| Wrong part brought | Full return visit plus a restock or dead stock | Serial number captured at booking, resolving mid-model variants |
| No part on the van | Return visit plus the customer's wait for shipping | Ordering-frequency report driving the standing van kit |
| Misdiagnosis on visit one | Return visit plus a credibility problem | Per-appliance service history showing the prior repair |
| Part in the other van | Return visit for stock you already own | Per-location inventory counts with recorded transfers |
| Customer not home / no access | Lost slot, rescheduled call | Confirmed appointment window with an automated ETA text |
| Repeat failure billed twice | Refund or an argument, sometimes a bad review | Warranty record tied to the specific unit and part |
The ordering loop has to close
When a part genuinely has to be ordered, the job is not finished — it is in a queue, and queues leak.
The loop that works: the technician records the part on the job while still on site, the order goes out as a purchase order against a supplier rather than as a text message someone might miss, the receiving step lands that part against the job it belongs to, and the return visit gets scheduled when the part arrives instead of when someone remembers. The customer hears from you at each of those steps without having to chase you.
The failure mode is a part that arrives and sits on the shelf for a week because nobody connected it to the customer waiting on it. That single gap produces most of the "I called three times and nobody called me back" reviews in this trade, and it is entirely a records problem. It is also the reason the diagnostic-fee conversation goes badly when it goes badly — a fee is easy to justify when the follow-through is visible, which is the practical argument in the diagnostic fee conversion piece.
What this asks of your software
Nothing exotic. It asks for four things: intake fields that capture model and serial and travel with the job; customer records that hold individual appliances with their own service history; real-time parts inventory that is tracked per van and decrements off the invoice; and purchase orders that connect an ordered part back to the job waiting on it.
Plenty of scheduling-first platforms handle the calendar well and stop at the parts. If you are comparing options for an appliance repair business, test the parts and history workflow specifically — see how the platform handles a second appliance in the same home, and whether it can tell you what is on van two. That is where the differences show up, and it is worth reading the honest comparison against Housecall Pro before committing.
First-visit fix rate is not a dashboard ornament. It is the compressed expression of whether your intake, your history, and your van stock are doing their jobs. Move it three points and you will feel it in the bank before you see it in a report.
Related reading: Appliance repair management software guide · Appliance repair dispatch and routing · Parts reorder alerts and purchase orders. For a complete machine-readable feature and pricing reference, see our LLM reference page.