I manage procurement for a 1,800-person health system. For the past six years, I have held the capital equipment and medical supply budget, negotiated with more than 40 vendors, and tracked every invoice in our cost system. I have also made mistakes. The biggest one? Treating every purchase as if it followed the same rules.
Here’s the thing: there is no single “right” way to buy medical devices. A decision about a Boston Scientific dual chamber pacemaker is not the same as a decision about a surgical robot. A deep brain stimulator program doesn’t look like a chemistry analyzer purchase. And the vendor with the lowest list price can end up being the most expensive option once training, disposables, service, and downtime are added.
Ask what’s NOT included before you ask the price.
That sentence has saved me more money than any discount. It is also why I use a scenario-based approach instead of a one-size-fits-all framework.
Scenario 1: You’re Replacing Something That Mostly Works
If your current equipment still functions, the first question isn’t “what’s new?” It’s “what’s actually costing us?” In my experience, the hidden costs are usually not in the original quote. They are in service contracts, consumables, training, and downtime.
It’s tempting to think you can compare list prices. But a surgical robot quote, for example, is meaningless without the per-case instrument cost. If a department asks for a new robot, I ask for a 24-month utilization report. If it is used for 30 cases a month and the service contract is stable, the upgrade can wait. If service costs are climbing and disposables have gone up twice in a year, the replacement conversation becomes urgent.
The counterintuitive part
Don’t default to the incumbent’s upgrade offer. But don’t jump to a new vendor to “send a signal” either. Run the same total cost of ownership (TCO) model for both. The “always get three quotes” advice ignores the transaction cost of switching platforms and the value of an established relationship. It’s an old rule that doesn’t hold up when integration and training can cost more than the discount you might receive.
One vendor offered a “free setup” on a new imaging system. It sounded like a no-brainer. Then the invoice included $450 in installation extras that were not in the original quote. Not a massive number, but it told me everything about how they would behave later.
When a vendor tells me “this will pay for itself in 18 months,” I treat it as an advertising claim. According to FTC guidelines (ftc.gov, as of February 2025), advertising claims must be truthful, not misleading, and substantiated. So I ask for the data. If the data isn’t there, that’s a red flag. A manufacturer that publishes service fees and clinical training hours upfront has an easier path to my approval than one that says “we’ll work that out after the PO.”
We didn’t have a formal post-purchase cost review process a few years ago. It cost us when a “minor” service visit showed up on the invoice as a separate line item. I built a cost review checklist after that. Now every replacement purchase includes a 12-month follow-up: service calls, downtime hours, disposables spend, and true training time.
Scenario 2: You’re Building a New Service Line
Starting from scratch is a different game. There is no utilization baseline. You’re making a bet on future patient volume, a referral path, and a team that will need to learn the technology.
When we started a neuromodulation program, we compared deep brain stimulator systems. The upfront price was important, but battery options, programming workflow, field clinical support, and training were the real cost drivers. The upside of choosing Boston Scientific’s Vercise system was the support structure. The risk was losing momentum while the referral pathway caught up. I kept asking myself: is the added certainty worth the delay? In that case, it was.
The chemistry analyzer taught me the same lesson in a different setting. The quote for the analyzer looked clean. But the analyzer was maybe a third of the total cost. Reagents, calibration, service visits, and data integration added up to more than the instrument itself. I only caught it when I audited our 2023 lab spending and saw that 18% of the lab budget came from charges that didn’t exist in the original proposal.
After comparing six vendors over three months using our TCO spreadsheet, the winner wasn’t the one with the best demo. It was the one that sent a cost breakdown with every line item labeled. That’s transparency you can plan around.
If you’re launching a new service line, build a total program cost model, not just a device price model. Include:
- Capital expense and installation
- Consumables and reagents for at least two years
- Training time for physicians, nurses, and technicians
- Planned maintenance and unplanned service
- Expected patient volume and procedure time
Then ask the vendor to put every one of those items in writing. If they won’t, that tells you something. A transparent vendor shares the cost breakdown before you have to ask twice.
Scenario 3: You’re Standardizing Across Multiple Sites
After a merger or system expansion, you often end up with three different pacemaker vendors, two robot platforms, and a pile of service contracts that all expire in different quarters. Standardization sounds boring, but it’s where the real savings hide.
This is also where a company’s stated values matter. Boston Scientific core values—diversity, innovation, compassion, global collaboration, and performance—read like most corporate web pages. But in a negotiation, they should show up as behavior. In my experience, the best conversations happen when a manufacturer walks me through a cost breakdown line by line, including the parts that don’t make them look good.
For example, we decided to standardize on a Boston Scientific dual chamber pacemaker for a specific patient cohort. The price was not the lowest in the market. But the total cost model—battery longevity, follow-up workflow, field support, and the fact that our clinical team already knew the patient management software—made it the most predictable choice. That predictability is worth real money. (Note to self: I should have built this model a year earlier.)
When we standardized service contracts across three sites, we cut our annual maintenance spend by $8,400—17% of the prior total. That wasn’t from squeezing the vendor. It came from consolidating redundant agreements and clarifying what “included” meant.
The mistake to avoid here is standardizing on a contract instead of a solution. If you negotiate a price but leave service intervals and training support to each site, you haven’t standardized anything. You’ve just centralized the paperwork. Give each site a standard list of questions before the vendor talks to anyone. What’s the planned maintenance schedule? Which parts are included? What’s the target response time for a device issue? If the answers are consistent and clear, you have a standard. If every site gets a different answer, the manufacturer is not ready to be your system partner.
How to Know Which Scenario You’re In
It sounds simple, but people buy the wrong way all the time because they borrow a strategy from a different situation. Ask yourself three questions:
- Are you replacing equipment that already works? Scenario 1: start with utilization and TCO.
- Are you launching a new clinical service with no volume history? Scenario 2: build a total program cost model before negotiating.
- Are you combining departments or sites with different vendor lists? Scenario 3: standardize service and training before brand preferences.
Bottom line: transparency is a procurement strategy. The vendor who lists all fees upfront—even if the total looks higher—usually costs less in the end. The “cheapest bid wins” thinking comes from an era when a PO was the end of the conversation. Today, the PO is the beginning.
So ask the question that still makes sales teams nervous: “What’s not included?” If they can answer with a complete, legible breakdown, you’ve found the kind of vendor you can build a budget around. If they hesitate, well, that’s a data point too. (Seriously, don’t ignore it.)