The interpretation rules treat goods put up in sets for retail sale as one article where the components are put up together to meet a particular need, are not normally sold separately in that combination, and come in retail packaging. A genuine activity kit can meet that test and take the heading of the component that gives it essential character, with the full set value entered under that heading. A warehouse bundle of unrelated bestsellers in one carton does not meet it, however the listing is written, and each item keeps its own heading, rate, and origin.
The file should record which treatment was applied and why, per kit, with the component list attached. Kits change: a component drops out, a promotional extra goes in, and the set analysis can change with the contents. A kit SKU whose contents drifted across a year of entries, all filed on the original analysis, is a quiet population of errors.
Even where a set is classified as one article, the cost model should start at component level: supplier cost per component, freight allocated by weight or cube as your tariff runs, then duty on the set treatment where it applies, or per component where it does not. Component level costing is what lets you answer the only bundle questions that matter: which component carries the margin, which one carries the duty, and what happens to the kit's price when one component's rate or origin changes under a new measure.
Mixed origin kits need the same discipline. Components from two countries in one retail set raise origin and marking questions that a single country of origin field in a spreadsheet will flatten into an error. Record origin per component, confirm the marking and entry treatment with your broker for the set as configured, and repeat the check whenever the sourcing of a component moves.
A limited time bundle assembled in the warehouse from stocked singles is not an imported set at all: the imports already happened, each on its own entry. Its landed cost is the sum of the components' landed costs plus the assembly and packaging cost, and the duty question is already settled. Confusing warehouse bundles with imported kits produces models that re duty the same goods on paper and price a promotion out of existence, or the reverse, and both errors survive because nobody reconciles the model to the entries.
Keep the two bundle types in separate tabs with separate rules, labelled by where the bundling physically happens. Most kit costing errors are really labelling errors about place, and place is a fact nobody should have to guess.
The customer sees one price and judges value against the sum of the parts sold separately. Your margin analysis should run the same comparison in reverse: kit contribution against the contribution of the same components sold singly, including the pick and pack saving of one shipment and the return complexity of a kit sent back in part. A kit that lifts conversion and basket size while holding contribution is a merchandising win. A kit that only moves the discount from the receipt total into the product cost is a relabelling, and the component build is how you can tell the difference.
No. Only combinations that meet the set conditions in the interpretation rules, put up together for a particular need in retail packaging, take a single classification by essential character. Other bundles are entered as their separate articles.
By the driver your freight actually charges on, usually weight or dimensional weight, applied consistently and documented. The allocation method matters less than using the real driver and keeping it stable so kit margins can be compared across periods.
No. The components were imported separately and entered on their own facts. The kit's cost is the sum of component landed costs plus assembly and packaging. Do not apply an imported set analysis to a domestic assembly.
The set analysis and the origin picture. Any content change should trigger a classification and cost review of the kit SKU before the new version ships, not after a quarter of entries has accumulated on the old analysis.