Parts, trucks, and first-time fix: HVAC inventory that stops the bleeding

Inventory feels like a back-office chore until you add up what it actually costs you: the mid-job supply-house runs, the callbacks for a part you didn’t have, the refrigerant that “walked off,” the money tied up in dead stock on a shelf. A shop with no real inventory system bleeds a meaningful chunk of its inventory value every year. Carrying costs alone run 20-30%, and write-offs, expediting fees, and stockout-driven lost jobs pile on top. This is the system that stops the bleeding, and it pays off most in the one number customers feel: first-time fix.

First-time fix is the metric that ties it all together

First-time-fix rate (FTFR), the share of calls resolved in one visit, is the operational heartbeat. Field-service benchmarks cluster around 70-80%; shops without an inventory system lag at 60-70% (roughly 1 in 3 calls needing a return), the target is 80%+, and top operators hit 82-92%. Every point of FTFR you gain is a callback you don’t eat, a customer who trusts you more, and a billable hour you keep. Most FTFR misses trace back to two things: misdiagnosis (see cutting callbacks) and not having the part on the truck. This guide is about the second.

The real cost of a single stockout

Put a number on the supply-house run so the whole team takes stocking seriously. One mid-job stockout costs roughly $350-$500 all-in:

• Windshield time to the supply house and back: ~$80-$120 (40-50 minutes of paid tech time).
• Fuel + vehicle wear: ~$25-$40.
• Supply-house markup over your distributor price: a 15-30% premium on the part.
• The lost/delayed billable call you had to push to make room for the trip: $150-$350 in revenue impact.

Do that a few times a week across a fleet and it’s tens of thousands a year, invisible on the P&L because it’s scattered across a dozen line items.

Stock trucks by actual usage, not a generic list

The biggest truck-stock mistake is loading every van with the same manifest. Build each van’s stock from that tech’s real last-12-months usage. A residential tech on 15-year-old split systems in one neighborhood needs a different kit than a commercial tech running rooftop units across office parks. Start from your high-velocity failure parts (capacitors, contactors, common motors, refrigerant, thermostats, igniters, fuses), then tune per tech and per the equipment mix in their territory.

Treat every van as a named inventory location. Parts loaded onto a van are a transfer; parts used on a job are a deduction tied to that work order. Without that structure, truck stock becomes a black hole: you can’t reorder accurately, you can’t see shrinkage, and you can’t trust any count.

Min/max reorder points, and flex them by season

Stop reordering by gut. Set a reorder point per SKU from real consumption:

Reorder point = (average daily demand × supplier lead time) + safety stock

Then flex it for HVAC’s brutal seasonality: high-velocity summer parts (capacitors, contactors, refrigerant) need min levels 30-50% higher heading into cooling season than their winter baseline (and heating parts the reverse). Review levels at least quarterly so you’re stocked ahead of the season, not scrambling during it. Two things make the formula actually work: run a 30-day parts-usage log before you set any min/max (gut-based stocking is the silent killer), and set lead-time buffers per SKU, not one generic number. HVAC parts lead times swing anywhere from 2 to 10+ days. Also cap each van’s stock value at roughly 1.5-2× its monthly usage. Anything beyond that is dead weight that just shrinks. Automate reordering where your software supports it, but only once your counts are trustworthy (below).

Shrinkage and accuracy: you can’t automate what you can’t trust

Two numbers gate everything:

• Inventory accuracy must stay above ~95% on high-frequency items before you can trust automated reordering. The manual-process industry average is a grim ~63%, meaning nearly 40% of parts are miscounted, mislocated, or don’t exist. Cycle-count high-velocity SKUs regularly instead of one dreaded annual count. Small shop (under ~5 trucks)? Don’t chase 95% across everything on day one. Start by cycle-counting your top ~20 SKUs and expand from there.
• The #1 accuracy killer is the verbal “I used a capacitor.” Require techs to scan or photograph every part as it comes off the truck at job close, tied to the work order. Memory-based logging is where counts (and billing) fall apart.
• Shrinkage hides in untracked truck stock, unbilled parts, and theft; hold it tight (aim for the low single digits, and treat anything approaching ~10% as a problem to hunt down). Call out warranty/“goodwill” parts as their own bucket: parts given away on no-charge calls are real shrinkage that otherwise hides. Put the shrinkage dollar figure on a monthly P&L line so you feel it, not just a percentage.

Use software, and connect it to the job

Manual spreadsheets break down past a couple of trucks. Field-service/inventory platforms (ServiceTitan, Simpro, Housecall Pro, and dedicated inventory tools) track parts across warehouse and vans, deduct on work-order completion, trigger reorder points, and (critically) make sure every part used gets billed to the job. Unbilled parts are pure margin leak and a shrinkage source. Connecting inventory to invoicing is where a system pays for itself twice: fewer stockouts and fewer give-away parts.

Checklist

• Measure your first-time-fix rate; target 80%+ and treat “didn’t have the part” misses as an inventory problem to fix.
• Put a dollar cost on a stockout (~$350-$500) so the team takes truck stock seriously.
• Build each van’s stock from that tech’s 12-month usage, not a generic list.
• Make every van a named inventory location; loads = transfers, uses = deductions on the work order.
• Run a 30-day usage log first, then set min/max per SKU (demand × lead time + safety stock) with per-SKU lead-time buffers; flex 30-50% by season; review quarterly.
• Scan/photograph every part at job close (no verbal logging); cap each van at 1.5-2× monthly usage.
• Hold inventory accuracy >95% (cycle-count high-velocity SKUs; small shops start with top ~20) before trusting auto-reorder.
• Drive shrinkage to low single digits; track warranty/goodwill giveaways separately; put the dollar figure on a monthly P&L line.
• Run it in inventory/field-service software that bills every used part to the job.

The bottom line

Inventory is a margin story disguised as a chore. Every supply-house run is $350-$500 and a callback risk; every mile of untracked truck stock is shrinkage; every unbilled part is give-away profit. Stock trucks by real usage, treat each van as a tracked location, set seasonal reorder points, hold your counts above 95% accurate, and let software bill every part. Do it and your first-time-fix rate climbs toward 90%, which your customers feel, your techs appreciate, and your P&L quietly rewards.

General information for HVAC business owners. Benchmark figures are industry ranges that vary by shop size, service mix, and region. Track your own usage and costs.