Most plants I talk to in MI, IN, OH, IL, and WI treat a tariff or resin spike like weather: they read it, they worry, they wait for the invoice.

That is backwards.

A 10% duty on a line that is 4% of purchased spend, already covered for 11 weeks, with a contract that resets in February, is not a 10% problem. It is a coverage problem. The plant that gets hurt is the one buying the same SKU from a single country, with 18 days of inventory, and a customer price that was locked last quarter.

What I built Costly to do is ugly and specific:

1. Take the SKUs you actually buy — not a news feed.

2. Attach HTS + origin only when we can stand behind it.

3. Convert the shock into dollars on your mix: best / expected / worst.

4. Rank the moves: pull forward, dual-source, substitute, or do nothing.

If the answer is do nothing, that is the product working. Most “cost intelligence” tools cannot say that because they get paid to look busy.

First categories we model: corrugate/paperboard, resin packaging, steel/aluminum, diesel/natural gas, selected food ingredients. If you run a 10–50 person shop in that band, the first useful output is a weekly action list — not a dashboard of red dots.

Legal line, because it matters: we do not classify goods for CBP, we do not file, and we do not give legal advice. We tell you where the dollars are.