Skip to main content

PRODUCT MATCHING

Home
>
Glossary
Glossary
>
Glossary

Definition

Product matching is the process of identifying and linking equivalent products between your catalog and those of your competitors. The goal is to ensure that when you compare prices, you are comparing the exact same product (same brand, same model, same packaging, same variant). Effective product matching is essential for reliable price monitoring.

Why it's important

  • Ensuring comparisons are meaningful: comparing the price of a 128 GB iPhone 15 with a 256 GB iPhone 15 makes no sense. Matching prevents these errors.
  • Automating monitoring at scale: without automated matching, it’s impossible to monitor thousands of products reliably and consistently.
  • Avoiding wrong decisions: Incorrect matching can lead you to lower your prices when comparing different products, unnecessarily eroding your profit margin.

A concrete example

An online retailer is selling the "Samsung Galaxy S24 Ultra 256 GB Black." At competitor A, the product is listed as "Galaxy S24U 256GB Black." At competitor B, it appears as "Samsung S24 Ultra 256GB Phantom Black." The matching algorithm uses EAN codes, product attributes (brand, model, capacity, color), and machine learning to identify that these three listings refer to the same product. Once the match is established, the system can compare prices reliably and automatically.

How to do it

Matching by EAN/GTIN: the most reliable method for manufactured goods. If two products have the same barcode, they are identical. Limitation: not all competitors display EANs.

Attribute matching: characteristics (brand, model, size, color, etc.) are compared. This requires standardizing the data (Samsung vs. Sam., 256 GB vs. 256GB).

AI-powered matching: Machine learning algorithms analyze product titles, descriptions, and images to identify matches even when no EAN is available. This is a powerful method, but it is more complex to implement.

Common Mistakes

  • Don't rely solely on the product title: titles are often incomplete or ambiguous. Cross-reference multiple attributes to confirm a match.
  • Ignore variants: A single product may exist in multiple versions (color, size, capacity). Matching must be performed at the variant level.
  • Don't overlook the packaging: a 6-pack is not the same as a single bottle, even if it's the same brand and product.

Learn more

  • Research & Data: Competitor price data to enable professional-grade product matching.
  • Solutions: Product matching and chaining to automate matching across your entire catalog.
  • Tip: Use operational pricing to organize your product catalogs and ensure reliable comparisons.
  • Resources: Check out our case studies to see how matching enhances competitive intelligence.

Mini FAQ

It depends on the industry. In electronics, where products are standardized, aim for 80 to 90%. In textiles or home decor, where products are more varied, 50 to 70% is already a good result.

Yes. Catalogs are updated as new products are added and existing ones are discontinued. Monthly or quarterly data synchronization ensures that the data remains up to date.

Almost. An occasional manual review is still necessary to validate uncertain matches and improve the algorithms.

‍