Commerce CensusContactRequest access

Home / Learn

Building price collection versus buying it

An honest comparison of running your own price collection against using an API, including the costs that only appear in month three.

Building looks cheap in week one. You write a parser, it works, and the whole thing seems like a solved problem. The costs arrive later and they are mostly not engineering costs.

The recurring ones: sellers change page structure without warning, so parsers rot continuously rather than breaking loudly. Access is rate-limited per network address, so volume needs address rotation, which is an ongoing bill and an ongoing operational surface. And matching across sellers is a permanent quality problem, not a feature you finish.

The one that surprises people is verification. If a value is wrong you usually cannot tell, because a wrong price looks exactly like a right one. Teams that build their own collection typically discover this after acting on bad data, not before.

Building genuinely makes sense in two cases: you need one seller only and it publishes an official interface, or your requirements are so specific that no general index would carry the fields you need. Outside those, the honest calculation is that you are choosing to operate a data-collection team as a permanent cost centre.

Keep reading

Guide

What is commerce intelligence?

Commerce intelligence is the layer above price monitoring: an index of what exists, who sells it, at what, and what that means.

Guide

Why percentile beats percent-off

A percentile against a product's own history is a defensible measure of discount depth; percent-off a reference price is not.

Guide

Product matching and why it fails quietly

How cross-seller product matching works, the three ways it fails, and why published match confidence matters.

Guide

GTIN, UPC, EAN and ASIN

How product identifiers relate, why check digits matter, and what to do when there is no identifier at all.

Guide

Data provenance in commerce data

Why every price should carry its origin and a confidence weight, and what goes wrong when it does not.