Every classification dispute in the world — a CBP ruling in New York, a BTI in Brussels, a classification notice in Hanoi — is argued through the same six General Interpretative Rules. If you understand them, you can predict how customs will decide.
GIR 1: heading texts and Notes rule everything
Classification is determined by the terms of the headings and the Section/Chapter Notes. Marketing names are irrelevant. Example: a "smart watch" is not classified as a watch of Chapter 91 by its name — Note 3 to Chapter 91 and the terms of heading 8517 must be examined to decide whether communication or timekeeping gives its character.
GIR 2: incomplete goods and mixtures
2(a): an unassembled bicycle shipped in a carton is still a bicycle (8712) if it has the essential character of the complete article. 2(b): references to a material include mixtures — the basis for moving to GIR 3 when mixtures create two candidate headings.
GIR 3: when two headings compete
3(a): the most specific description wins — "electric shavers" (8510) beats "electro-mechanical domestic appliances" (8509). 3(b): sets and composite goods follow the component giving essential character — a gift set of pasta, sauce and grater is classified by the pasta. 3(c): if still tied, the heading last in numerical order wins.
GIR 4, 5 and 6
GIR 4 (goods most akin) is rarely used — it is a signal your description is too thin. GIR 5 covers cases and packing: a violin case entered with the violin follows the violin; packing "clearly suitable for repetitive use" does not. GIR 6 repeats the whole method at subheading level: compare only subheadings of the same level, with their Subheading Notes.
Practical takeaway
Write your product description in GIR language: what it IS, what it is made of, what gives it its essential character. That is exactly the reasoning path our AI follows — and the format a customs officer expects to read.