Marketplace Sold Comps Analyzer

Analyze user-provided completed marketplace sale records into sold-comps pricing guidance. The Actor does not autonomously discover listings, scrape marketplaces, use proxies, require logins, or call paid external APIs.

What It Does

This Actor groups completed sale records by item and turns them into comparable-sales metrics: comp count, recent sales count, sold-price percentiles, spread, outliers, marketplace and condition distribution, confidence score, sample quality, and a recommended listing price range.

Who It Is For

Resellers, marketplace operators, pricing analysts, liquidation teams, and ecommerce teams that already have completed sale exports and need a repeatable way to convert raw sold records into listing-price decisions.

Input

Provide records with completed sales. Each valid record needs an item or title, a sold price, and a sale date. Recommended fields are item, soldPrice, soldAt, marketplace, condition, saleType, currency, and url. groupKey, priceField, dateField, and dedupeKey let you map your own export columns.

Output

Each dataset row is one item group with completed-sale metrics: compCount, recentSalesCount, medianSoldPrice, p25SoldPrice, p75SoldPrice, recommendedListPriceLow, recommendedListPriceHigh, confidenceScore, sampleQuality, marketplace and condition distributions, recommended action, score explanation, and source record indexes.

How It Works

The Actor validates that each record describes a completed sale by requiring an item key, numeric sold price, and parseable sold date. It normalizes item names, prices, dates, currency, marketplace, condition, and sale type, then deduplicates by marketplace, item, currency, sold price, sold date, and source key.

For each item group it calculates min, 25th percentile, median, 75th percentile, mean, max, spread percentage, and IQR-based outlier count. recentSalesCount counts sales within 90 days of the latest sale date in that item group. soldVelocityProxy is recent sales per 30 days inside the observed window; it is only a proxy from the supplied records, not a marketplace-wide demand estimate.

confidenceScore ranges from 0 to 100. More comps add coverage confidence, more recent sales add recency confidence, a tight spread improves confidence, very wide spread lowers it, and outliers lower it. sampleQuality is strong at 75+, usable at 55+, and thin below 55. Recommended action follows sample quality: list within the recommended range, review conditions before listing, or collect more sold comps.

Example

{
  "records": [
    {
      "item": "Vintage Camera X",
      "soldPrice": "$100.00",
      "soldAt": "2026-05-01",
      "marketplace": "eBay",
      "condition": "used",
      "saleType": "auction",
      "url": "https://m.example/1"
    },
    {
      "item": "Vintage Camera X",
      "soldPrice": 140,
      "soldAt": "2026-07-01",
      "marketplace": "Mercari",
      "condition": "excellent",
      "saleType": "buy_it_now",
      "url": "https://m.example/2"
    }
  ],
  "groupKey": "item",
  "priceField": "soldPrice",
  "dateField": "soldAt",
  "dedupeKey": "url"
}

Example output row:

{
  "itemKey": "vintage camera x",
  "compCount": 2,
  "recentSalesCount": 2,
  "medianSoldPrice": 120,
  "recommendedListPriceLow": 110,
  "recommendedListPriceHigh": 130,
  "confidenceScore": 53,
  "sampleQuality": "thin",
  "recommendedAction": "collect_more_sold_comps"
}

Pricing

PAY_PER_EVENT event analysis_record_emitted is charged once per emitted sold-comps analysis row. Invalid records, duplicates, skipped rows, and charge-limited rows are not billable.

Limitations

This Actor only analyzes completed sale records supplied by the user. It does not scrape eBay, Amazon, Facebook Marketplace, or other marketplaces; it does not verify whether a sale is real; it does not perform FX conversion; it does not forecast future demand; and it does not use live marketplace data unless you provide it as input.

Integration

Run it directly with JSON input, from an Apify task, or downstream from another Actor/export that produces completed sale records in a default dataset. Results are written to the default dataset and the run summary is written to the OUTPUT key-value store record.

Differentiation

Use Marketplace Sold Comps Analyzer when you have completed sale records with dates and need sold-comps pricing guidance. Use Ecommerce Price List Analyzer when you have active offer or catalog records and need current price-list spread and outlier analysis.