Ramp vs Brex for Matching Late Freight Bills to a PO

• The original purchase order amount, tagged with a PO number before anything else gets added to it
• Customs and duty charges, which typically arrive two to four weeks after the container clears
• Drayage and inland freight, billed separately from ocean freight and often by a different vendor
• Branch-level local freight, paid ad hoc by whichever buyer needed a delivery expedited

Each row needs the same PO number attached, and the worksheet isn't complete, meaning the landed cost isn't real, until every late-arriving charge has posted against it. A PO marked closed the day the container clears is closed on inventory but still open on cost.

A purchasing team that closes a PO the moment goods are received, then reconciles landed cost weeks later once customs and drayage bills arrive, is running two separate processes that have to reconnect after the fact. If the person closing the PO isn't the same person who later matches the customs bill, that reconnection depends on a shared reference number existing in both places, which is exactly the kind of detail that gets dropped under deadline pressure.

Keeping the PO open, or flagged as pending final cost, until landed cost is fully reconciled rather than closing it at receipt removes the need for anyone to reconnect the two processes manually later.

A branch buyer expediting a local delivery on a personal or general-purpose card is solving the immediate problem, getting the shipment moving, while creating the exact same tracing problem that customs charges create at the container level, just smaller and more frequent. A dozen small branch freight charges across a month, each on a different card and none tagged to a PO, adds up to a real distortion in landed cost that nobody catches because no single charge looks significant on its own.

Requiring a PO number on branch freight charges the same way it's required on the original purchase order closes this gap without adding much friction to a buyer trying to solve a delivery problem quickly.

Distributors carrying inventory on a revolving credit line, rather than paying cash for every purchase order, are typically borrowing against a rate tied to the prime rate, and it's worth knowing that number before treating a card's short payment float as a substitute for that financing¹. A card program is built for tracking and control, not for carrying inventory cost over weeks or months the way a credit line is designed to.

Confusing the two, using card float as informal financing because it's convenient, tends to surface as a surprise when the statement comes due and the inventory hasn't sold through yet.

A distributor running a handful of branches gets the most from Ramp's fast card issuance for branch buyers and its custom field support for requiring a PO number before a freight charge posts. Its automated matching also reduces the manual work of tying a dozen small branch charges back to the right purchase orders each month.

A distributor running many branches, or one importing from multiple countries with customs and duty charges in different currencies, benefits from Brex's multi-currency features and should ask directly how it handles a multi-branch structure. The tradeoff is the same one that shows up across most comparisons: expect a fuller application at onboarding, since a platform built for a multi-branch, multi-currency structure generally asks to see more financial history before opening that up. That's less of a burden for an established distributor than for one just setting up its first formal card program.

The single most common source of bad landed-cost data is a purchasing system that marks a PO closed the moment inventory is received, which is accurate for inventory tracking but wrong for cost tracking since customs, drayage, and duty charges haven't arrived yet. Reports pulled from a PO system in that state understate true landed cost on every recent purchase, sometimes by a meaningful margin once freight surcharges are included.

Flagging a PO as cost-pending rather than fully closed until every expected charge has posted keeps reports honest, even if it means carrying a few dozen POs in a pending state at any given time.