The data-project trap
The most expensive sentence in supply chain is “first we need to get the data right.” It postpones every negotiation behind an integration project measured in quarters, and when the warehouse finally lands, the contracts it was meant to inform have rolled over — escalators intact. Meanwhile the categories kept renewing on the calendar, which does not wait for clean data.
The trap rests on a false premise: that category-level decisions need line-item perfection. They do not. Deciding whether exam gloves deserve a competitive bid requires knowing roughly what you spend, what you pay per unit on the items that matter, and when the contract ends. It does not require five years of normalized transaction history across every category you own. Precision is worth buying — for the category you are about to negotiate. Everywhere else, ranges are enough to rank the work.
What a usable baseline actually is
For one category, a baseline you can negotiate from — and defend to finance later — is three things: twelve months of spend, current unit prices on the highest-volume items, and the contract’s terms and end date. In most hospitals that is an afternoon with accounts payable, the item master, and the contract file. Not an integration. Not a project charter.
Write the baseline down and agree it with finance before the negotiation starts. This is the step that makes every later claim auditable: the “before” was fixed while it was still just a fact, not ammunition. A baseline reconstructed after the deal closes will always be suspected of flattery, and the suspicion is fair.
- Twelve months of category spend — accounts payable already has it
- Unit prices on the top items by volume, from the item master or recent invoices
- Contract terms, end date, and notice window from the contract file
Price and utilization are different levers
One discipline separates credible savings math from theater: never blend what a unit costs with how many units you use. Negotiation moves price. Standardization, par-level discipline, and clinical practice move utilization. Different levers, different owners, different timelines — and if a reported result mixes them, finance cannot tell whether procurement negotiated well or the census dropped.
This is also why patient data has no place in this work. Category economics run on purchase orders, invoices, and contracts. A platform architected so that no patient data is ever required — Zero-PHI by design, not by policy — takes the longest item off the critical path: the security review that would otherwise precede the first useful conversation.