Tech for Appliance Repair: The Software Stack That Actually Makes Money

Your tech is under a dishwasher 10 hours a day, the part's on order with no tracking, and the warranty claim is aging at 60 days. The right stack fixes all three. The stack in order, what each piece costs, and when to buy versus build. One weekend of setup ends the text-message dispatch era — if you buy in the right order. Also see Appliance Repair, AI for Appliance Repair, and Payments for Appliance Repair.

The stack, in the order that matters

Buy in this order — each layer pays for the next:

• 1. Dispatch and scheduling — the operating system. Routes, calendar, dispatch, invoices, payments in one app. Half-day route packing is the feature that matters here.
• 2. The phone system — the front door. A business line that rings the right person, records calls, texts customers. Your personal cell is not a phone system.
• 3. Accounting — the scoreboard. QuickBooks Online or Xero, synced to dispatch. Midnight bookkeeping ends when data stops being retyped.
• 4. Parts lookup — the accuracy layer. Model-to-part lookup with photos (so the right control board gets ordered the first time), tied to the job. A mis-ordered $210 board costs a second trip and a week of waiting.
• 5. Photo documentation — the proof layer. CompanyCam (Core $63/mo for one user, Crew $119/mo for three, annual billing): model-plate photos, old-part photos, before-and-after — GPS-pinned to the job address. The model-plate photo is your warranty-claim evidence.
• 6. Marketing presence — the lead pipe. Google Business Profile (no cost to claim), a simple website, Google Local Services Ads — in that order. A $10,000 website before 50 reviews is jewelry on a work truck.

The platforms, honestly compared

Prices checked October 2026; annual billing shown first.

• Rossware ServiceDesk — $150/mo; the longtime appliance-and-electronics repair staple, with SmartParts model-to-part lookup (about $20/mo) built into its DNA and warranty-billing workflows the generic platforms don't have. Best for: shops that live on parts ordering and warranty claims. Skip if: you need modern mobile-first dispatch — it's a Windows-era workhorse with a learning curve.
• Housecall Pro — Basic $59/mo ($79 monthly, 1 user); Essentials $149 ($189, 5 users); MAX $299 ($329, 8 users). Cards from 2.59%. Best for: 1-to-5-tech shops wanting everything in one modern app, with Spanish support. Skip if: warranty-claim volume is your core — the warranty workflows are thin.
• Jobber — Core $49/mo ($29 annual, 1 user) to Plus $699/mo for 15 users; cards 2.9% + $0.30. Best for: quote-heavy shops wanting deep QuickBooks sync. Skip if: you want everything included.
• Workiz — no-cost tier (2 users), paid from ~$200/mo. Best for: the one-tech shop that needs dispatch today and a budget of zero. Skip if: you outgrow two users — the jump to paid is steep.

Two notes the brochures skip. Rossware's SmartParts lookup is the feature no generic platform replicates: one-click part identification with photos from the model number, which is why appliance shops tolerate the dated interface — mis-orders drop to near zero. And if warranty work is real volume for you, the platform decision is really a warranty-workflow decision; the dispatch features are table stakes. Whatever you pick, run 20 real jobs through the trial before you port the number — migrations are where shops go to die.

Buy vs. build: when custom software makes sense

For 98% of shops, the answer is buy. This trade has a purpose-built platform (Rossware) and solid generics — there is almost nothing to build.

Build when the software is the advantage:

• Warranty-claim automation across multiple warranty companies — if you file 200 claims a month across five companies with five portals, a unified claim compiler is a real build.
• A customer portal for property managers: every unit's appliance history, work approvals, and invoices — the thing that wins the 200-unit contract.
• Parts-inventory intelligence — which parts to stock on the van by failure-rate data, if your platform's inventory is a list with no opinions.

A real build starts at $25,000–$75,000 plus 15–20% a year in maintenance. Nearshore senior engineers cut that roughly in half — Nova offers a first consultation free, and shops engage dedicated 2–3 person senior-engineer teams through that consultation when a real build is justified. If you cannot name the revenue it unlocks, you do not need it.

Most "custom" needs die at the integration layer: Rossware or Housecall Pro plus Zapier or Make covers automated follow-ups, review requests, parts-arrival texts, and QuickBooks sync for under $200/mo. Build only when the software wins the contract — and if you can't name the contract, you don't need the build.

What the whole stack costs: 1 tech, 3 techs, 8 techs

Monthly totals, annual billing, before card fees:

• 1 tech: Workiz no-cost tier + QuickBooks Simple Start $38 + business line ~$20 = ~$60/mo. Or Housecall Pro Basic $59 instead of Workiz = ~$120/mo.
• 3 techs: Housecall Pro Essentials $149 + QuickBooks Essentials $85 + phone ~$79 + AI answering ~$79 + CompanyCam Core $63 = ~$455/mo.
• 8 techs: Rossware $150 + SmartParts $20 + QuickBooks Essentials $85 + phone ~$199 + AI answering ~$150 + CompanyCam Crew $119 = ~$725/mo.

Software should run 1–2% of revenue — a $600k shop at $455/month is at 0.9%. The expensive mistake is never the subscription; it is the shop running on texts, memory, and a warranty claim aging at 60 days.

Marketing spend sits outside the stack: Google Local Services Ads run about $40–$70 per appliance-repair lead in 2026 benchmarks, and you can dispute invalid leads for credit. At a ~50% book rate that's roughly $100 per booked job — know the number before you raise the weekly budget.

Where shops waste money

• ServiceTitan at 2 techs — a reporting suite for the office manager you do not have.
• Six apps that do not talk — every handoff between them is a dropped ball.
• Seats nobody uses — audit twice a year; the $29/mo seat for the tech who quit in March adds up.
• Parts inventory by gut feel — $5,000 of dead stock on the shelf while the $210 board you actually need is on order. Let the data pick the stock.
• "No-cost" software with strings — if the price is zero, your data is the product or locked in. Check the export story first.
• Building custom too early — a $40,000 portal for a 3-tech shop is a hobby, not an investment.
• The AI add-on nobody turned on — dispatch and review apps now upsell AI features that bill monthly while nobody configures them. Audit: if no one's trained it, cancel it.
• Marketing software before the reviews exist — Local Services Ads with 12 reviews is paying to advertise that nobody trusts you yet.

FAQ

Rossware or Housecall Pro? Parts-and-warranty volume: Rossware — the SmartParts lookup and warranty workflows are the decision. Modern dispatch and ease of use: Housecall Pro. Run 20 real jobs through each trial first.

When does ServiceTitan make sense? Around 10+ techs with an office manager and real reporting needs. Before that, a Ferrari for grocery runs.

Do I need to leave my personal cell? Yes. Call recording ends every "he said the price was X" dispute, and AI answers while you're under a dishwasher.

iPads or phones for techs? The phones they already carry, in a rugged case. A separate tablet is one more thing to charge, lose, and never update.

When do I hire an office manager instead of more software? Software first, until ~5–8 techs. Hire the office manager when dispatch, warranty claims, books, and follow-ups exceed one person's day.

Should I stock more parts on the van? Stock what the failure data says, not what your gut says. The right $210 board on the van is a same-day repair; the wrong $5,000 of stock is a shelf of regrets.

This month: pick the dispatch platform, port the number, connect QuickBooks, turn on parts-arrival texts and the diagnostic-fee rule. One weekend ends the text-message dispatch era. AI answering, custom warranty automation, fancy reporting — all after the foundation. Then measure: unbilled diagnoses, parts-rebooking rate, warranty-claim age. The stack is working when all three move. When money moves faster, read Payments for Appliance Repair so processors do not take their cut.