I still remember the exact moment my spreadsheet betrayed me.
It was a Tuesday in Q2 2024. I was sitting in my office at a regional hospital group—we run 4 imaging centers across two states—staring at a comparison of two CT scanner quotes. One from a familiar vendor, one from a name we hadn't worked with before. The numbers were clear on paper. The decision should have been easy.
It wasn't. And the rabbit hole I went down over the next three months fundamentally changed how I think about procurement in medical imaging.
The Setup: A $1.4M Procurement That Looked Like a No-Brainer
Let me back up. I manage the medical equipment budget for our group—about $3.2 million annually for imaging hardware and service contracts. In early 2024, we needed to replace a CT scanner at our busiest imaging center. The existing unit was an older GE model, and our service costs had crept up. Time to upgrade.
The clinical team wanted a 128-slice system. Our radiologists had a preference list: GE Healthcare, then Siemens, then Philips. No surprises there. GE had the incumbent advantage—our service team knew the platform, we had spare parts inventory, and the radiologists liked the workflow.
I went through our standard process: get three quotes, compare on price, service terms, and training. Standard stuff. Or so I thought.
Vendor A (not naming names, but it was one of the big three): quoted $847,000 for the hardware, $72,000/year for a 5-year service contract, and $18,000 for installation and clinical training. Total 5-year TCO: $1,225,000.
Vendor B (a smaller but established player): quoted much lower—$689,000 for hardware. But the service contract was $89,000/year after the first year, and training was separate. When I ran the 5-year numbers, it came out to roughly the same: $1,218,000. Almost identical. That was suspicious.
Vendor C: a new entrant to the US market, offering aggressive pricing. Their quote: $602,000 for hardware, $48,000/year for service. Installation included. 5-year TCO: $842,000.
I went back and forth between Vendor C and the established options for two weeks. A nearly $400,000 savings over 5 years? That's not nothing—that's an entire additional MRI upgrade budget.
But something gnawed at me.
The Turn: When You Realize the Spreadsheet Is Lying
I've managed procurement long enough to know that sticker price is rarely the full story. But even I missed something.
In my initial TCO model, I accounted for hardware, service, installation, and training. What I didn't fully account for was the operational friction of switching platforms.
We'd been a GE house for 13 years. Our PACS was integrated with GE's ecosystem. Our technologists had been trained on GE consoles. Our radiologists had customized protocols and hanging protocols for GE images. Our service engineers had GE certs. Our parts bin was full of GE-compatible consumables.
Switching to Vendor C meant:
- New training: 3 days per technologist for the new console. At 12 FTEs, that's 36 days of lost productivity—about $43,000 in labor cost alone.
- PACS integration: Our IT team estimated 2 weeks of custom integration work. Plus the RIS system needed updates.
- Protocol migration: Each radiologist had 3-5 custom protocols for different exam types. Rebuilding those on a new platform would take months. In the meantime, throughput would drop.
- Service overlap: We'd need to maintain GE service contracts on our other scanners for parts and support during the transition. No discounts for partial fleet.
- Radiologist preference: I don't want to understate this. If the docs aren't comfortable with the images, no amount of spreadsheet savings matters.
I built a second spreadsheet. A real one. And that's when the numbers went sideways.
When I added the switching costs—training, integration, protocol migration, productivity loss, and dual-service overlap for 6 months—the gap between Vendor C and the GE option narrowed to $127,000 over 5 years. Not nothing, but way less than $400,000.
And that's assuming everything went smoothly. It never does.
The Decision: Going With What We Knew
So I went back to GE Healthcare—but not with my tail between my legs. With data.
I sat down with their account team and showed them the competitive quote. Not as a threat, but as a question: "How do you justify the premium?"
To their credit, they didn't give me a sales pitch. They walked me through the TCO breakdown I hadn't fully appreciated until I did my second spreadsheet:
- Service response time: Their guaranteed 4-hour response vs. Vendor C's 8-hour. For a CT scanner doing 40+ exams per day, four hours of downtime is $12,000 in lost revenue. That alone was worth a premium.
- Parts availability: We had a local GE parts depot. Vendor C's closest depot was in another state. More downtime risk.
- Software updates: GE's service contract included all software updates. Vendor C's contract required a separate upgrade fee for major software versions.
We negotiated. I'm not gonna lie—I'm proud of what we got. GE gave us a 7% discount on the hardware, threw in an extra year of service, and included remote monitoring at no charge. The final 5-year TCO came to $1,098,000.
Versus Vendor C's best-case $842,000, yes, GE was still $256,000 more expensive. But when I added the real switching costs—conservatively estimated at $150,000 in productivity loss and integration—the actual gap was about $106,000 over 5 years.
For that $106,000, we got:
- Zero training cost for existing staff
- No PACS integration headaches
- No protocol migration
- Same service team that already knew our site
- Parts compatibility across our existing fleet
- Radiologists who could read immediately at full efficiency
That felt like a good trade.
The Lesson: TCO Isn't Just Numbers—It's Pain Avoidance
Here's what I learned, and what I now tell every procurement manager I mentor: The biggest cost in medical imaging equipment isn't the hardware. It's the transition.
I've managed our imaging budget for 6 years now, tracking every invoice, every service call, every hour of downtime. In that time, I've seen procurement teams save $300,000 on a scanner only to lose $200,000 in hidden switching costs—training, integration delays, protocol rebuilding, radiologist frustration, and lost throughput.
The industry has evolved significantly since 2020. Back then, "buy the cheapest that meets specs" was common advice. But as of 2025, that advice is outdated—because the cost of disruption has gotten more expensive. Staff are harder to train. Integration is more complex. Every hour of downtime costs more as patient volumes grow.
Does that mean you should always stick with your incumbent vendor? No. But it does mean you need to calculate the real cost of switching, not just compare hardware quotes.
I have mixed feelings about the whole experience. Part of me wishes I'd been braver and gone with the new vendor. Another part knows that the GE decision was the safer bet for our specific situation—and we got better terms because we competed the bid properly.
If you ask me, that's the real skill: not picking the cheapest vendor or the safest vendor, but understanding the total cost landscape well enough to make an informed trade-off.
The scanner went live in September 2024. Three months later, we hit our target throughput. Zero unplanned downtime in the first 6 months. The radiologists didn't miss a beat.
Bottom line: sometimes the most expensive option on paper is the cheapest one in practice. And a good procurement manager knows how to tell the difference.
