Product Matching

Product matching is the process of deciding which competitor listing genuinely corresponds to which of your products, and most price monitoring problems are matching problems. Price Trakker matches competitor listings to your catalogue using identifiers where they exist, product attributes such as brand, model, specification and pack size where they do not, and direct listing URLs where the competing page is already known. Automation carries the volume and people verify the ambiguous cases before they reach a comparison, and corrections persist so accuracy improves rather than resetting on every run. Why an incorrect match matters: a wrong match creates a confident but false price gap. Matched to a single unit when the competitor sells a multipack, you cut a price that was never undercut; matched to a premium variant, you hold a price that is genuinely uncompetitive. Matching large catalogues: at tens of thousands of products across several competitors and channels, matching only works if automation handles volume and human verification handles the cases automation cannot settle, and because that work sits with us the catalogue size is not your team’s problem. Catalogues without GTIN, EAN, UPC or MPN: own-brand, trade, industrial and long-tail ranges frequently carry no identifiers, so brand, model, specification, pack size and direct listing URLs are used instead. Variants, bundles and difficult matches: size and colour variants, pack sizes and multibuys, bundles including an accessory yours does not, refurbished listings and near-identical model numbers are the recurring hard cases. Variants and pack sizes are treated as distinct products rather than merged, and non-comparable listings are flagged for review rather than forced into a comparison. FAQ. Can you monitor competitor prices without GTINs? Yes, using attribute matching and direct listing URLs. Is product matching automated or done by people? Both; automation at scale with human verification on ambiguous matches. Why does an incorrect product match matter? It produces a false price gap that looks real and destroys trust in the dataset. How are bundles and multipacks handled? As distinct products, flagged for review rather than compared with a single unit.