Pricing and commerce
Product data
Information that describes products: identifiers, titles, attributes, images, prices, availability and reviews.
In practice
The raw material of pricing, content and catalogue decisions.
Separate product, variant and offer
Product data describes identity and attributes, while an offer adds seller-specific price and purchase conditions. A variant may change colour, size or capacity without changing the product family. Keeping these levels separate helps avoid overwriting one seller’s price with another or attaching availability for one size to every size.
Example: modelling a catalogue item
An illustrative record for a bottle might include brand, model, capacity, material and identifier. A retailer offer adds seller, price, currency, availability, URL and observation time. A two-bottle pack needs its own quantity context. A flat table can still work, but its key must identify the level being stored so repeated observations are not mistaken for duplicates.
Normalise without losing evidence
Use consistent units and controlled values where appropriate, while retaining the original source value when review requires it. Missing identifiers should remain distinguishable from invented ones. Product titles and images can assist matching, but they do not automatically prove two offers are identical. A documented schema makes the data easier to join, compare and validate across channels.
Further reading
Aperture tracks prices, promotions and availability with AI product matching and dated evidence, and Import.io feeds deliver the same data to your own systems.
Aperture pricing intelligence →