Skip to content

Latest commit

 

History

History
125 lines (99 loc) · 8.09 KB

File metadata and controls

125 lines (99 loc) · 8.09 KB

Price discovery

A single "price" is a fiction. A procurement-grade workflow tracks at minimum eight price tiers and synthesizes them into a fair-price band (aspirational / target / walk-away) before the recommendation step.

The eight price tiers

Tier What it is Where to find
MSRP / RRP / List Manufacturer's suggested retail. Anchor, not binding. Often inflated to make discounts look bigger. Manufacturer product page
Street price Actual transaction price across authorized retail. Always below MSRP for mature SKUs. Regional aggregators (Google Shopping, PriceSpy, Idealo, Kakaku, regional equivalents — see step 2 table)
Net price Out-of-pocket after coupons, cashback, loyalty, bundled rebates. The decision anchor. Seller page after applying discounts
Refurbished / open-box / B-stock Manufacturer- or retailer-restored. Open-box ~10–30% off new; refurb usually ≥90-day warranty; B-stock cosmetic-grade. Manufacturer outlet, Best Buy Outlet, Amazon Renewed, retailer refurb
Used market eBay sold listings, specialist resellers. Truth-tier for resale value. eBay sold (&LH_Sold=1&LH_Complete=1), MPB/KEH (cameras), specialist forums
Gray market / parallel import Genuine product, unauthorized channel. 10–25% cheaper but voids warranty in most regions; region-locked firmware risk. Cross-border marketplaces
Bundle price SKU sold with accessories/extended warranty. Standalone price often matched, bundle "discounted". Retailer listing
Trade-in residual OEM credit for old unit. Anchors TCO and replacement cycle. OEM trade-in program

The 3-quote rule

Institutional procurement minimum: 3 independent suppliers/channels per SKU before committing. For consumer buying, 3–7 is the practical range before admin burden exceeds informational gain. Always run at least 3 channels.

Price-discovery sub-process

For each surviving candidate from step 5, run in parallel where possible:

1. MSRP capture

WebFetch the manufacturer product page. Record list price + any active OEM promo.

2. Aggregator scan

Use the user's region (from step 4) to pick the right aggregators. Pull top 3 lowest sellers from each.

Region Primary aggregators Major first-party marketplaces
US Google Shopping, PCPartPicker (for PC), Amazon, Newegg Amazon, B&H, Adorama, Newegg, Best Buy
UK PriceSpy, PriceHippo, Google Shopping Amazon UK, Currys, John Lewis, Argos
EU (general) Idealo (DE/AT/IT/FR/ES), Geizhals (DE/AT), PriceSpy (Nordics) Amazon DE/FR/IT/ES, MediaMarkt, Saturn, Otto, Fnac
Canada PCPartPicker, Newegg.ca, Google Shopping Amazon.ca, Best Buy, Canada Computers
Australia StaticICE, PCPartPicker AU, Shopbot Amazon.au, JB Hi-Fi, Officeworks
Japan Kakaku.com, Price.com Amazon.co.jp, Rakuten, Yodobashi
India Compare India, Mysmartprice, PriceDekho Amazon.in, Flipkart, Croma
Turkey Akakce, Cimri Amazon.com.tr, Hepsiburada, Trendyol, N11
Other / unlisted region WebSearch "<category> price comparison <region>" to identify local aggregators Local marketplaces + Amazon if available

General caveat: aggregators surface listing price, not transacted price, and may not strip coupons / loyalty discounts. Manual spot-check at the seller page is mandatory in every region.

3. Price history

  • Amazon SKUs: camelcamelcamel.com/product/<ASIN> (free, full history) or keepa.com (11 marketplaces, deeper data).
  • Capture: lowest-ever, 90-day low, 12-month median.
  • Use to detect fake "was/now" discounts — the FTC bans inflating "was" prices under 16 CFR 233.1. If price-history shows the "was" never held for >2 weeks in 6 months, the discount is fictional.

4. Used / refurb floor

  • eBay sold listings: append &LH_Sold=1&LH_Complete=1 to any search URL. Use median of 5–10 sales. Ignore Best Offer strikethroughs — actual sale price is hidden.
  • Manufacturer certified-refurb outlet (Apple, Sony, DJI, Best Buy Outlet) — usually 10–25% off new with ≥90-day warranty.
  • Domain-specific: KEH / MPB for cameras; Bicycle Bluebook for bikes; eBay for everything else.

5. Peer-tier benchmark

Repeat steps 2–3 at lower depth for 2–3 peer SKUs from the candidate slate. Output: peer-tier median price. Used to anchor whether the SKU itself is overpriced for its tier, independent of any one retailer's markup.

6. Cross-border check (optional)

Only if user signaled openness to importing.

  • WebFetch the listing on a major cross-border retailer (Amazon US/DE/UK, B&H, etc.) priced in the source currency.
  • Apply landed-cost math for the user's region: CIF (cost + insurance + freight) + customs duty + VAT/sales tax + any excise / luxury tax + shipping.
  • Use the user-region's official customs calculator (search "<country> import duty calculator" and prefer the government source). Examples:
    • US: USITC Tariff Database; for personal imports the de minimis threshold varies per country of origin.
    • EU (any member state): TARIC database (ec.europa.eu/taxation_customs/dds2/taric).
    • UK: gov.uk Trade Tariff tool.
    • Turkey: ticaret.gov.tr.
    • Other regions: search for the equivalent national customs authority tool.
  • Verify de minimis rules in user's region — many countries have removed or lowered the low-value customs exemption in 2024–2026; assume full duty + VAT applies unless verified otherwise.
  • Rule of thumb: an item must be roughly ≥30–40% cheaper abroad before import math breaks even for personal-volume buying. Exact break-even depends on the destination's duty + VAT stack.

7. Red-flag pass

For each top-3 seller in step 2, scan for:

  • Third-party marketplace seller (vs. seller-fulfilled-by-marketplace vs. first-party retailer)
  • Listing age < 30 days (new sellers = unknown reputation)
  • "Was/now" discount where price-history shows the "was" never held
  • Refurb listing with no warranty term or < 90-day warranty
  • FX-margin manipulation on local listings of imported goods (cross-check against the user-region central bank's effective rate)
  • Generic stock images instead of actual product photos
  • Seller rating < 95% or < 100 reviews on first-tier marketplaces

8. Synthesize fair-price band

Three numbers per candidate:

  • Aspirational = min(12-month low, eBay sold median × 1.15) — best plausible deal seen recently
  • Target = peer-tier median − 5% — the "happy buy" price
  • Walk-away = min(peer-tier median + 10%, MSRP) — ceiling; above this, kill the deal

9. Deliverable

Per candidate, a structured object:

{
  sku: "Sony WH-1000XM5",
  region: "<user region>",
  currency: "<user currency>",
  msrp: <amount>,
  current_low: <amount> (<aggregator>, <seller>),
  12mo_low: <amount> (camelcamelcamel or Keepa, FX-converted to user currency),
  used_median: <amount> (eBay sold, age <N>-months-old units),
  fair_band: { aspirational: <amount>, target: <amount>, walk_away: <amount> },
  red_flags: ["<list of any red flags from step 7>"],
  best_channel: "<retailer name + authorized status>",
  return_policy: "<window + restocking fee + shipping>",
  authorized_dealer: <true/false>
}

This object feeds into the TCO computation (step 7) using target as the assumed purchase price, and into the report's Price Discovery table.

Key sources

Cross-region:

  • camelcamelcamel — Amazon price history (US/UK/DE/FR/IT/ES/CA/JP/AU/MX), free, full
  • Keepa — 11 Amazon marketplaces, paid for deep history
  • PCPartPicker — PC components with native price history (US/UK/DE/CA/AU/FR/IT/ES + others)
  • eBay sold listings — used-market truth in any region; append &LH_Sold=1&LH_Complete=1
  • MPB, KEH — used camera gear with public price indices (US/EU/UK ship globally)
  • Manufacturer direct — sometimes cheaper than authorized retailer; always check first

Regional aggregators (see step 2 table for full list per region). For any region not pre-mapped, derive via WebSearch "<category> price comparison site <country>".

Citation rule

Price data in the report: ≥1 aggregator link + ≥1 primary seller link, per candidate.