Trust
Methodology
How we collect, clean, and present resale market data.
The Data Pipeline
Every price you see on ResalePriceGuide comes from real marketplace listings — never from estimates, predictions, or fabricated numbers. Our pipeline ensures that each data point is sourced, cleaned, and verifiable before it appears on the site.
1. Raw Listing Collection
We pull active and recently sold listings from five marketplaces: eBay, StockX, GOAT, Grailed, and Mercari. Each listing capture includes the asking or sale price, product size, condition, platform, and timestamp.
2. Size Filtering
Listings are filtered to a consistent size standard (adult men's sizing unless otherwise noted). This prevents infant, toddler, or women's-specific sizes from skewing the adult resale price distribution.
3. Deduplication
Cross-posted listings (the same pair listed on multiple platforms by the same seller) are identified and counted once. Duplicate listings inflate apparent supply and distort median pricing, so they are removed before aggregation.
4. Outlier Removal
Statistical outliers — listings priced far outside the normal distribution (e.g., a $2,000 listing for a shoe trading at $200, often a typo or bundle listing) — are removed. The number of raw listings collected and the number removed as outliers is documented on each report.
5. Price Calculation
From the cleaned dataset, we calculate the median price (the middle value, less sensitive to extremes than the mean), the P25–P75 price range (the interquartile range showing where the middle 50% of listings sit), and the premium vs. retail percentage.
6. Demand Signal
Demand signal is a directional summary combining listing volume, price trend, and sell-through patterns. It ranges from "hot" through "warm," "neutral," "cool," and "cold." It is a qualitative read, not a precise metric.
Point-in-Time Snapshots
Our data reflects point-in-time snapshots, not real-time prices. Each product guide and market report is stamped with a "last data update" date so you know exactly when the snapshot was taken. Data is refreshed periodically — we prioritize accuracy over speed, which is why we do not show live ticking prices.
Between snapshots, prices may move. Always check the "last data update" date on any report before making decisions based on the data shown.
No Fabricated Numbers
We never fabricate, estimate, or model prices when real listing data is unavailable. If a product does not have sufficient cleaned data to produce reliable metrics, we mark it "data pending" rather than showing a placeholder number.
This means you may encounter gaps in coverage. We consider this a feature, not a bug — a missing data point is more honest than an invented one.
Platform Breakdown
Each report includes a platform breakdown showing how many listings were analyzed from each marketplace (eBay, StockX, GOAT, Grailed, Mercari) and what share of the total each represents. This transparency lets you assess the breadth of data behind every number.
Brand Resale Index
The Brand Resale Index is a transparent composite score measuring overall secondary-market strength for a brand. It combines four components, each derived from validated MarketSnapshot data with approved provenance:
1. Value Retention (40%)
For each qualifying product, value retention = current median price ÷ original retail price × 100. Brand-level score is the median of qualifying product values (robust to outliers). A score of 100 means current resale median equals original retail price. This is the only component where 100 has that specific meaning.
2. Price Momentum (25%)
Based on validated 30-day median price change. Score = 100 + median 30D percentage change across qualifying products. 100 means no net change. If 30D history is unavailable for a brand's products, momentum is marked as pending — no fake history is substituted.
3. Liquidity / Market Activity (20%)
Measures relative secondary-market activity. Per product: cleaned listing count × √(marketplace count). Brand-level score is average activity per qualifying product, normalized to 100 = market median. This ensures brands with more tracked products don't dominate simply because they have more coverage.
4. Price Stability (15%)
Measures pricing consistency. Per product: 1 − (P75 − P25) ÷ median (clamped to 0–1, higher = more stable). Brand-level score is the median of qualifying product values, normalized to 100 = market median. Brands are not penalized for having different absolute price levels — only for dispersion.
Overall Index
The overall Brand Resale Index is a weighted average of available component scores. If a component lacks sufficient data (e.g., momentum when no 30D history exists), its weight is redistributed across available components. A score of 100 represents baseline neutral resale-market performance — above 100 indicates stronger-than-baseline, below 100 indicates weaker. 100 does NOT mean resale equals retail. That interpretation applies only to the Value Retention component.
Qualification Requirements
- Minimum 2 qualifying products per brand with current validated snapshots
- Minimum 10 cleaned observations per product
- Minimum 1 marketplace represented per product
- Snapshots must be within 14 days of the current date
- Approved provenance only (synthetic_test excluded)
- Real known retail price (not estimated)
Brands that do not meet these requirements are shown as "Index pending — additional market data required" rather than receiving a fake neutral score.
Category Awareness
The index supports category-specific calculations (Sneakers, Apparel) in addition to an overall index. Brands across unrelated categories are not directly compared without normalization. Liquidity and Stability scores are normalized relative to the market median within the selected category.
Calculation Versioning
Each index calculation is tagged with a calculation version (e.g., BRI_V2_1). If methodology changes in the future, stored historical snapshots retain their original version, ensuring historical consistency.
Limitations
- The index measures market performance, not causation.
- Active listing volume does not equal completed-sale demand.
- Component availability varies — a brand may have Value Retention and Liquidity data without Momentum history.
- Historical brand index changes will become available as BrandIndexSnapshot records accumulate over time.