Annual tariff budgets built on one blended rate die by spring in a fast moving year. Reforecasting works product by product: current official rates on planned volumes, list changes in the pipeline, and sourcing decisions already made. The result is a range with owners, not a single wrong number.
Why blended rates fail under movement
A blended rate averages products whose duties never move with products under active review. When one category's rate doubles, the blend moves a little and the category's margin collapses a lot. Leadership sees a small variance while a product line quietly goes underwater.
For more on this, see this related guide and our earlier explainer on the same theme.
Blends also hide timing. A rate announced mid year may apply only to goods shipped after a date, hitting some purchase orders and not earlier ones. Only an order level view captures that, and budgets are spent at order level whether the forecast admits it or not.
Building the reforecast bottom up
Start with open and planned purchase orders by product and origin. Apply the official rate in force for each expected entry date, including stacking duties that apply together. Add announced changes with their effective dates as scenarios, not as certainties buried in the base case.
Show three numbers: base with current law, downside with announced changes applied, and upside if exclusions or sourcing shifts land. Ranges invite planning. A single point in a volatile year invites false confidence and angry surprises.
Assigning variance owners
Every material variance line should carry a name: who is watching this rate, this exclusion expiry, this sourcing move. Tariff variance reviewed without owners becomes weather reporting. With owners, each review meeting ends with decisions: reprice, resourse, redesign, or consciously absorb.
Readers also found this companion piece from our archive useful.
Track forecast accuracy as its own metric. If the tariff forecast misses in the same direction repeatedly, the process has a bias, usually optimism about exclusions or delay. Naming the bias is the first step to pricing it in.
Communicating the reforecast upward
Lead with exposure concentration: the few products and origins that drive most of the swing. Executives do not need four hundred lines, they need to know where the money is and what levers exist, with dates by which each lever must be pulled.
Close with what would change the range: pending announcements, exclusion decisions, shipment timing choices. A reforecast that states its own triggers for the next update teaches the business how tariff risk actually moves, which improves every future budget conversation.
Frequently asked questions
How often should a tariff budget be reforecast in a volatile year?
Quarterly at minimum, and immediately when a change hits a top exposure product. In calm years annual budgets survive. In moving years, quarterly is the difference between steering and reporting wrecks.
What is duty stacking?
Several duties applying to the same goods at once, for example a base rate plus a special action rate plus a metal derivative duty. Forecasts that apply only one rate understate cost badly on stacked products.
Should announced but not final rates enter the base forecast?
Keep the base on law in force, and carry announced changes as a clearly labeled downside scenario. Mixing them into the base without labels makes the forecast unreadable when announcements change or slip.
Which products drive most tariff variance?
Usually a small set: high volume, high rate, or actively reviewed categories. Finding that set is the first reforecast task, because attention and levers should concentrate there.