How to Calculate Refurbishment Costs: Unit Economics That Actually Hold Up

Most refurbishment operators calculate "cost" as: what I paid for the unit plus what I spent on parts. That model consistently under-counts by 30–50%, which is why so many operations look profitable on paper and then wonder why margins compress at scale. This guide covers the complete cost stack — every component that belongs in your per-unit economics — and shows how to model it before you bid on a lot, not after you've already processed it.

The Seven Cost Components in a Complete Refurbishment Model

A rigorous per-unit cost model has seven components. Missing any one of them distorts your margin calculation:

1. Acquisition cost — what you paid per unit including buyer's premium and lot-level allocation
2. Inbound logistics — freight, pallet fees, receiving labor
3. Labor: testing and diagnosis — time × hourly rate for the full test sequence
4. Parts and materials — components replaced, cleaning supplies, packaging
5. Labor: repair and cosmetic work — time × rate for actual refurbishment
6. Overhead allocation — facility, equipment amortization, utilities, management, software — allocated per unit
7. Return risk reserve — a per-unit buffer that reflects your expected return rate and return cost

Components 1–5 are what most operators track. Components 6–7 are what most operators miss, and they're often the difference between a 35% gross margin and a 22% realized margin.

Industry Benchmarks by Category and Condition Grade

The following benchmarks reflect typical operator data in the US and EU recommerce market for consumer electronics in 2025–2026. Use these as starting points and replace with your actual tracked data as you accumulate it.

The wide range in C-grade smartphones and B-grade laptops reflects the variance in what actually needs repair — a C-grade phone with a cracked screen (screen replacement: $30–$60 for most models) vs. one with just cosmetic scratches (parts: near $0) are categorized identically but have radically different process costs. This is why condition assessment before committing to refurbishment is essential.

Overhead Allocation: The Component Most Operators Miss

Overhead is the sum of all fixed and semi-fixed costs divided by units processed per month. For a typical small-to-mid refurbisher:

• Facility: $1,500–$4,000/month for 500–2,000 sq ft of refurb space
• Diagnostic equipment amortization: $200–$500/month (assumes $5,000–$12,000 in equipment amortized over 24 months)
• Software subscriptions: $100–$400/month (ERP, listing tools, diagnostic software)
• Management and administrative: $2,000–$8,000/month depending on scale
• Utilities: $200–$600/month

If your total overhead is $5,000/month and you process 800 units, your overhead allocation is $6.25 per unit. At 200 units per month, it's $25 per unit — a significant difference that explains why low-volume operations often have worse unit economics than they expect.

The overhead tipping point: Most refurbishers see a step-change improvement in per-unit economics somewhere between 400–600 units/month, when fixed costs spread across enough volume that per-unit overhead drops below $10. Below that threshold, per-unit economics often look better than they are if overhead isn't fully allocated.

Return Risk Reserve: Pricing for Realistic Outcomes

Every unit you sell carries some probability of return. A typical return costs:

• Return shipping: $6–$15 (your cost)
• Receiving and retest: $4–$8 in labor
• Relisting markdown: 10–20% below original price
• Total return cost: ~$25–$45 for a $100–$150 unit

At a 5% return rate on B-grade phones with an average $130 sale price and $35 return cost, your return reserve should be: 5% × $35 = $1.75 per unit. This seems small, but it prevents you from accidentally treating returns as a surprise cost rather than a predictable operational expense. For C-grade with 10–12% return rates, the reserve climbs to $3.50–$4.20 per unit.

Worked Example: iPhone 13 B-Grade Lot from B-Stock

You bid on a 50-unit B-Stock lot described as "customer returns, mixed models, iPhone 12 and 13 generation." Manifest shows 30 iPhone 13 and 20 iPhone 12. Your winning bid: $3,200 ($64/unit). Here's the full cost model:

19% gross margin is acceptable for B-grade phones if throughput is high. But notice: if the operator had only counted acquisition + parts ($64 + $8.40 = $72.40), they'd have projected a 48% margin. The full cost model is 29 points lower than the incomplete model — which is exactly the kind of margin illusion that causes refurbishers to over-bid on procurement and then wonder why profit doesn't materialize.

The Lot-Level P&L Model

Per-unit cost models assume you know the condition mix before you buy. You often don't — especially on unsorted returns lots. The lot-level P&L model accounts for outcome uncertainty by modeling a grade distribution assumption and testing break-even at different bid prices.

For the same iPhone 13/12 lot example, assume your grade distribution estimate from comparable B-Stock lots is:

• A-grade outcome: 20% of units → sell at $155 (iPhone 13), $85 (iPhone 12)
• B-grade outcome: 45% of units → sell at $140 (iPhone 13), $75 (iPhone 12)
• C-grade outcome: 25% of units → sell at $105 (iPhone 13), $55 (iPhone 12)
• Parts/scrap: 10% of units → sell at $15–$25

Weighted average revenue per unit ≈ (0.20 × $155) + (0.45 × $140) + (0.25 × $105) + (0.10 × $20) = $31 + $63 + $26.25 + $2 = $122.25 (iPhone 13 units).

Weighted process cost (excluding acquisition) for iPhone 13: ~$42–$50 depending on outcome mix (C-grades and parts cost more to process, A-grades cost less).

Maximum bid price to hit a 25% gross margin: $122.25 × 0.75 − $46 = $45.69 per unit. At 30 iPhone 13 units, maximum bid contribution from this model segment: ~$1,370. Add the iPhone 12 segment calculation and you have a data-driven maximum bid for the lot.

This is the model professional liquidation buyers run before every bid. It's not a perfect prediction — actual condition distribution will vary from the estimate — but it establishes a margin floor that protects you from the single biggest procurement mistake: winning a lot that looks cheap and discovering it doesn't cover costs at any sale price you can realistically achieve.

How Process Costs Vary by Volume

The per-unit cost stack isn't static. Three components scale with volume in ways that compound:

• Testing labor efficiency: Experienced inspectors on a known product category run 15–25% faster than their first month. A 12-minute test becomes 9 minutes — saving $1.80/unit at $18/hr. Across 500 units/month, that's $900.
• Parts cost: Buying battery kits in quantities of 50+ typically drops unit cost 20–35% vs. single purchases. Screen replacement parts follow similar volume curves.
• Overhead allocation: At 300 units/month, overhead is often $15–$20/unit. At 1,000 units/month in the same facility, it drops to $5–$7. This is the dominant factor in why refurbishment economics improve with scale.

The practical implication: your cost model at month 3 of operation will be materially different from month 12. Revisit and recalibrate quarterly, particularly the overhead allocation rate and labor efficiency assumptions.

Where Operators Most Often Go Wrong

• Using "refurbishment fee as % of value" as the model: This approach is intuitive but wrong. A $300 iPhone 12 and a $300 MacBook Air have radically different actual process costs. Per-unit absolute cost models are more accurate.
• Not counting inbound logistics: B-Stock and Liquidation.com lots often carry $80–$200 in freight. On a 30-unit lot, that's $3–$7 per unit — significant against a $25–$35 margin target.
• Ignoring marketplace fees in the cost model: eBay's 12.5% + PayPal/Managed Payments fees on a $140 sale is $17.50. Amazon's referral + FBA fees on a $140 sale can reach $35–$45. These belong in the cost model, not as a post-margin surprise.
• Treating D-grade/parts units as zero cost: Parts units cost the same to receive, test, and disposition as sellable units. If 10% of your lot is parts and you're selling them at $20 each, you're still spending $8–$12 in process cost to handle them. That's net $8–$12 per parts unit — which should reduce your max bid for the lot.

Building Your Cost Model: What to Track From Day One

You can start with a spreadsheet. Track these fields per unit: source lot, SKU, condition on intake, grade after refurb, labor time (testing, repair), parts used and cost, final sale price, channel, marketplace fee, return (Y/N), return cost if applicable. After 60–90 days, you'll have enough data to calculate your actual average costs by category and grade — and those numbers will be more accurate than any benchmark.

The goal is a cost model that you can run in 60 seconds on any lot before bidding. Not because the estimate will be perfect, but because even a ±20% accurate model is far better than the alternative: discovering your margins after the fact.

To see how Recyscope automates per-unit cost tracking and margin scoring at the lot level, visit the Refurbishment Operations pillar or request early access.