A liquidation lot closes in 22 minutes. The manifest shows 85 units of mixed smartphones from a major retailer. Your instinct says this looks good — the category is strong, you know the brand. But instinct is not a bid strategy. Every profitable liquidation buyer has a repeatable evaluation framework they run before every bid, regardless of time pressure. This guide covers that framework in full.
The Five-Layer Pre-Bid Evaluation
Lot evaluation is not a single calculation — it is five sequential filters. Failing any filter is a reason to pass, or at minimum to apply a steep discount to your maximum bid.
Layer 1: Operational Fit
Before any financial analysis: can you actually process this lot? Check your current queue size against weekly processing capacity. A lot that pencils at 35% gross margin is worthless if your processing backlog means it sits untouched for 3 weeks while you are paying for the capital deployed. The rule: do not bid on a lot you cannot process within 5 business days of receipt unless you have intentionally structured a "buy and hold" position with the holding cost modeled in.
Layer 2: Manifest Quality Assessment
The manifest is your primary source of condition information. Not all manifests are equal. Before trusting any condition distribution estimate, assess the manifest quality:
| Manifest Type | What It Contains | Reliability | Adjustment to Revenue Estimate |
|---|---|---|---|
| Itemized with IMEI/serial + condition per unit | Each unit listed individually with grade and sometimes MSRP | High (85-95% accurate) | Use stated distribution, apply 5% downside buffer |
| Category-level manifest (model + grade counts) | Lists models, quantities by grade, no per-unit detail | Medium (65-80% accurate) | Shift distribution 10-15% toward lower grades |
| Summary only (lot description with % estimates) | "Approximately 60% B-grade, 40% C-grade" type description | Low (45-65% accurate) | Apply 20-25% discount to expected revenue |
| No manifest / "as-is, untested" | Category and approximate quantity only | Very low / unknown | Assume worst-case grade mix; bid at extreme discount or skip |
Layer 3: Market Price Research
For each model in the lot, you need current realized prices (not listed prices) by condition grade. The tools for this:
- eBay Sold Listings (last 30 days): Filter by sold, then by condition label (Good, Very Good, Excellent). This shows actual transaction prices, not aspirational asking prices. For a lot evaluation, use the median of the last 20-30 sold listings, not the high end.
- Keepa (for Amazon): Check the used/refurbished price history on the product's Keepa chart. Are prices stable, rising, or declining? Is there a seasonal pattern? How many sellers are active?
- B-Stock historical bid data: If you have purchase history on the platform, what did similar lots from this storefront sell for in the last 90 days? This tells you both market pricing and competitive bidding levels.
For a mixed-model lot (multiple SKUs), you need to weight the revenue estimate by the estimated unit distribution. Do not apply a single average price to the whole lot — a lot with 20 iPhone 14 Pro units and 65 iPhone 11 units is not valued at the average of the two models' prices.
Layer 4: Full Cost Stack Calculation
The most common error in lot evaluation is using a partial cost model. The full cost stack for any liquidation lot includes seven components:
| Cost Component | Typical Range (per unit) | Notes |
|---|---|---|
| Acquisition (bid price + buyer's premium) | Variable | Add platform buyer's premium (10-18%) to winning bid |
| Inbound freight | $1.50–$5.00 | Per unit; varies by lot size, distance, and shipping method |
| Processing labor (intake + test + grade) | $4–$12 | Varies by category complexity and team efficiency |
| Repair parts (if applicable) | $0–$35+ | Estimate based on expected C-grade volume and repair profile |
| Listing + photography labor | $2–$6 | Lower with listing tool automation; higher for detailed condition photos |
| Marketplace fees (sale) | 8–15% of sale price | eBay ~12.5-13.5%, Amazon Renewed ~15%+, direct channel lower |
| Return risk reserve | $3–$10 | Expected return rate × average return cost; 5% rate × $25/return = $1.25/unit |
Operators who model only acquisition + parts typically overstate margins by 25-40 percentage points. A lot that looks like 45% gross margin at acquisition + parts becomes 22-28% when the full cost stack is applied.
Layer 5: Qualitative Risk Factors
Even a lot that passes the financial analysis can fail qualitative checks that should adjust or eliminate the bid:
- Source retailer reliability: Some platforms have storefronts with histories of manifests that significantly overstate grade quality. Track your lot outcomes by storefront — if a storefront's actual grade distribution consistently underperforms the manifest, apply a larger haircut to that storefront's stated grades.
- Activation lock / Find My Device exposure: For smartphones and tablets from consumer return sources, activation lock affects 3-8% of units on average. If a lot comes from a source known for consumer returns (rather than corporate/B2B), build this into your cost model: 5% activation lock rate × acquisition cost = dead unit write-off.
- Category timing risk: Is there a new model release imminent in this category? A flagship smartphone lot purchased one month before the successor releases will face 15-25% price compression within 30 days. Check product release calendars before bidding on any current-generation flagship category.
- Lot size vs. channel capacity: A 200-unit lot of the same iPhone model may compress eBay pricing if you list them all simultaneously. Large single-SKU lots should factor in the market impact of your own supply — either list in smaller batches over time or route excess to bulk channels.
The Max Bid Calculation
Once all five layers are evaluated, the max bid formula is:
Max Bid per Unit = (Weighted Avg Revenue × [1 - Target Gross Margin]) - (All Non-Acquisition Costs per Unit)
Total Max Bid = Max Bid per Unit × Estimated Unit Count
Adjusted Max Bid = Total Max Bid ÷ (1 + Buyer's Premium Rate)
Worked example — 85-unit mixed smartphone lot, Balanced Growth mode (25% gross margin target):
- Estimated grade distribution after manifest haircut: 30% A-grade, 45% B-grade, 20% C-grade, 5% non-functional
- Weighted average revenue: (0.30 × $195) + (0.45 × $145) + (0.20 × $88) + (0.05 × $15) = $58.50 + $65.25 + $17.60 + $0.75 = $142.10
- Non-acquisition costs per unit: freight $3.20 + processing $8.00 + parts $6.50 + listing $3.00 + marketplace fees 12.5% ($17.76) + return reserve $5.00 = $43.46
- Max Bid per Unit = ($142.10 × 0.75) − $43.46 = $106.58 − $43.46 = $63.12
- Total Max Bid = $63.12 × 85 = $5,365
- Adjusted for 13% buyer's premium: $5,365 ÷ 1.13 = $4,748 is your maximum winning bid
Bid above $4,748 and your gross margin falls below 25%. If you want a margin buffer (recommended for manifests with medium reliability), reduce by 10-15% to $4,035-4,274.
Building Your Lot Evaluation Log
The single most valuable investment for improving lot evaluation accuracy over time is a structured log of every lot you bid on — won or lost. Record:
- Date, platform, storefront, lot ID
- Your max bid and winning bid (if you won)
- Estimated grade distribution at bid time
- Actual grade distribution after processing (if won)
- Actual revenue per unit vs. projected
- Reason for pass if you passed (over budget, qualitative risk, capacity)
After 30-50 lots, this log reveals systematic patterns: which storefronts consistently deliver better or worse than manifest, which categories you consistently over-estimate or under-estimate revenue for, and what your actual margin delivery rate is vs. your targets. This data is more valuable than any pricing tool.
For the cost calculation methodology, see How to Calculate Refurbishment Costs. For market research methods, see Market Analysis Best Practices. For platform-specific evaluation notes, see the Platform Overview guide.
Apply Your Bid Framework to Every Lot, Automatically
Recyscope builds your margin thresholds and cost model into the lot evaluation — so every bid decision is consistent, documented, and data-backed regardless of time pressure.
Request Early Access