Kepner-Tregoe Vendor Scoring for Medical Device Automation: Why the Lowest Bid Almost Always Costs More

Kepner-Tregoe vendor scoring is a structured decision analysis method that ranks medical device automation vendors by separating mandatory pass/fail qualifications from weighted preference criteria, then verifying scores across engineering, management, and operations before selection.
Kepner-Tregoe vendor scoring gives medical diagnostic manufacturers a structured, weighted method to evaluate automation vendors beyond price. It separates mandatory qualifications (MUSTs) from weighted preferences (WANTs), forces cross-functional scoring verification, and exposes vendors who underbid by hiding capability gaps. Picking the lowest bidder without this process consistently leads to change orders, schedule overruns, and validation rework that dwarfs the original price difference.
You’ve got three bids sitting on your desk for a new automation procurement project. One is $200K less than the other two. Management sees the number and wants to move. Engineering sees the requirements response and wants to pump the brakes.
This is where projects go sideways. Not in commissioning. Not during validation. Right here, at vendor selection, when someone decides that price is the only score that matters.
I’ve run 96+ automation projects across 25+ years in medical diagnostic manufacturing. I’ve watched the lowest bidder win against engineering’s recommendation and come in a year and a half late, way over budget, buried in change orders. I’ve also watched the right vendor, picked through a disciplined Kepner-Tregoe scoring process, deliver on time and on spec. The difference isn’t luck. It’s method.
Kepner-Tregoe decision analysis gives you that method. It doesn’t eliminate judgment. It structures it so every stakeholder’s priorities are visible, weighted, and verified before anyone signs a purchase order.
What Is Kepner-Tregoe Vendor Scoring and How Does It Apply to Automation Procurement?
Kepner-Tregoe (KT) vendor scoring is a decision analysis framework that separates mandatory qualifications from weighted preferences, then scores each vendor against both. It was developed by Charles Kepner and Benjamin Tregoe in the 1960s and has been used across regulated industries for decades. The Kepner-Tregoe methodology breaks every vendor decision into two distinct layers.
MUSTs are pass/fail gates. Binary. Either the vendor meets the requirement or they’re eliminated. No amount of low pricing rescues a vendor that fails a MUST. Examples in medical device automation include ISO 13485 certification, documented GAMP experience, and financial stability above a defined threshold.
WANTs are weighted preferences. Each gets a score from 1 to 10 based on importance, and each vendor gets rated against every WANT. Multiply the rating by the weight, sum the totals, and you’ve got a ranked list that reflects actual project priorities instead of gut feel.
That’s the structural difference between KT and just picking whoever comes in cheapest. The scoring happens before emotions, politics, or budget pressure enter the room.
Why Do Standard Bid Comparisons Fail in Medical Device Automation?
A bid comparison spreadsheet tells you one thing well. Price. It tells you nothing about whether the vendor actually understood your requirements, whether they’ve built GAMP-compliant systems before, or whether they’ll staff your project with their A-team or hand it to a junior crew after the contract is signed.
Standard bid comparisons fail because they treat automation like a commodity purchase. It isn’t. A fill-and-finish line for IVD cartridges isn’t interchangeable the way office supplies are. The integration work, the validation documentation, the FAT/SAT execution, the IQ/OQ/PQ support, the change order management, all of that lives outside the base bid number.
Here’s what happens when you skip structured scoring. Three vendors bid. Management picks the lowest. Engineering flags concerns about the vendor’s requirements response. They didn’t address half the URS properly. But the price gap is significant, and the budget conversation wins.
Six months in, change orders start stacking. The vendor quoted capabilities they didn’t have. They subcontract work they said they’d handle in-house. Timelines slip. Validation documentation arrives incomplete. The project that was supposed to save $200K ends up costing $400K more than the second-lowest bidder would have. And it’s a year and a half late.
I’m not describing a hypothetical. That’s a real project.

How Do You Build a Kepner-Tregoe Scoring Matrix for Automation Vendors?
Building a KT matrix for automation vendor selection follows a specific sequence. Skip a step and the scoring breaks down, because the whole framework depends on getting MUSTs right before WANTs enter the picture.
Step 1. Define Your MUSTs
MUSTs are non-negotiable. They’re the baseline qualifications every vendor has to meet just to stay in the conversation. In medical diagnostic manufacturing, your MUSTs typically include items like these, though every project will have its own specific requirements.
- ISO 13485 certification or equivalent quality management system documentation
- Demonstrated GAMP validation experience on regulated production equipment
- Financial stability verified through auditable records, D&B reports, or equivalent
- Proper response to the User Requirements Specification (URS), addressing every line item, not just the easy ones
- Adequate staffing capacity for the project timeline without overcommitting resources across other jobs
Any vendor that fails a single MUST gets eliminated. Period. That’s the entire point. A vendor with the lowest price but no GAMP documentation experience doesn’t advance to the WANT scoring, because they’ll create validation nightmares downstream that cost multiples of whatever you saved on the bid.
Step 2. Define and Weight Your WANTs
WANTs are where the real prioritization happens. And this is also where internal politics surface, which is actually a good thing when it happens before vendor selection instead of during the project.
Each WANT gets a weight from 1 to 10. Here’s where the conversation gets interesting, because different stakeholders weight differently.
- Management tends to weight financial criteria highest. Total cost, payment terms, delivery schedule impact on cash flow.
- Engineering weights technical capability. Did they respond to the requirements document properly? Do they have the right controls experience? What’s their track record on similar builds?
- Operations and maintenance care about long-term support. Spare parts availability, training programs, remote diagnostics capability, response time commitments.
- Quality cares about validation deliverables. GAMP documentation completeness, prior FDA audit history, change control procedures.
The weighting conversation forces alignment. Or at least forces disagreement out into the open where you can deal with it, instead of discovering it six months into a project that’s already hemorrhaging money.
Step 3. Score and Verify
Score each vendor against each WANT on a scale, typically 1 to 10. Multiply each score by the WANT’s weight. Sum the weighted scores for a total. The math is straightforward.
The critical step that separates a useful KT process from a checkbox exercise is verification. You don’t let one person score in isolation. The automation engineer builds the initial scoring. Then you verify those scores with management, operations, maintenance, and quality. Everyone sees the same matrix. Everyone can challenge a score.
This verification step is where I’ve seen projects saved. Engineering scores a vendor low on requirements compliance. Management scores them high on cost. The conversation happens before the PO is issued, not after the vendor is on site and failing to deliver.
What Criteria Belong in a Medical Device Automation Vendor Scorecard?
The specific criteria depend on your project, your product, and your regulatory environment. But across 96+ projects in IVD, consumable diagnostics, and lab component manufacturing, certain categories show up on every scorecard that produces a good outcome. Here’s what a typical scoring matrix covers for vendor qualification in medical device automation.
| Category | MUST or WANT | What You’re Really Evaluating |
|---|---|---|
| Quality system certification | MUST | ISO 13485 or equivalent. No cert, no conversation. |
| GAMP documentation capability | MUST | Can they produce IQ/OQ/PQ protocols, trace requirements through testing, and deliver a complete validation package? |
| Financial stability | MUST | Will they still be in business for warranty support in 3 years? Verifiable through D&B or audited financials. |
| URS response completeness | MUST | Did they address every requirement? Incomplete responses signal capability gaps they’re hoping you won’t notice. |
| Industry reputation and references | WANT (8-10) | Talk to their last 3 clients. Ask specifically about change orders, timeline adherence, and documentation quality. |
| Technical approach and design capability | WANT (7-9) | Is their proposed approach sound, or are they planning to figure it out as they go? Review their design methodology. |
| Project management structure | WANT (6-8) | Dedicated PM, clear communication cadence, documented escalation path. Not a salesperson doubling as your project contact. |
| Change order history | WANT (7-9) | How many change orders on their last 5 projects? A vendor with consistently high change order counts is either under-scoping or over-promising. |
| Training and post-installation support | WANT (5-7) | What’s the handoff plan? Who trains your operators? What’s the support response time after warranty? |
| Total cost of ownership | WANT (8-10) | Beyond the bid number. Spare parts pricing, maintenance contracts, software licensing, future upgrade paths. |
Notice the weighting suggestions on the WANTs aren’t fixed. That’s intentional. Your project drives the weights. A company running their first GAMP-validated automation line might weight documentation capability at a 10. A company with a mature quality system might weight it at a 6 and push technical approach higher. The scorecard adapts. The discipline doesn’t.
What Actually Happens When Management Picks the Lowest Bidder?
I’ll tell you exactly what happens, because I’ve lived it.
Management won on the lowest cost. Gave the cheapest vendor 100% on their financial score. The more expensive vendor, the one engineering preferred? 50%. The engineering team flagged that the winning vendor didn’t respond to the requirements document properly. Scored them accordingly. But the financial weighting swamped the technical scores.
The project came in about a year and a half late. Way over budget. Change orders piled up because the vendor had quoted capabilities they didn’t actually possess. Some automation companies, and I hate to say it, lie about their skill levels. They’ll tell you they’ve done GAMP work before. They’ll tell you they understand your process. Then they get on site and it becomes obvious they’re learning on your dime.
The change orders alone exceeded the original price difference between the two vendors. By a lot. Add the schedule delay, the production revenue lost during that extra 18 months, the validation rework, the internal engineering hours burned managing a vendor who was in over their head. The “savings” from choosing the lowest bid turned into the single largest cost driver on the entire project.
This isn’t an unusual outcome. It’s the predictable outcome when cost scoring dominates a KT matrix without adequate weight on technical capability and requirements compliance.

How Does Proper Kepner-Tregoe Scoring Prevent Vendor Selection Failures?
The KT framework doesn’t eliminate disagreement between management and engineering. It surfaces it. And it surfaces it at the right time, before the purchase order goes out, not after the vendor is on your production floor.
Three mechanisms make the difference.
MUSTs eliminate unqualified vendors before cost enters the picture. If a vendor can’t meet your baseline qualifications, their price is irrelevant. You don’t score a vendor who failed a MUST. They’re out. This prevents the scenario where a low bid from an unqualified vendor creates pressure to override engineering judgment.
Weighted WANTs force stakeholders to declare what matters. When management says cost is a 10 and engineering says requirements compliance is a 10, both priorities are visible in the matrix. You can run the scores both ways and see the impact. What happens if we weight cost at 10 and compliance at 6? What happens if we flip those? The math shows you the tradeoff clearly. No guessing, no backroom lobbying.
Cross-functional verification catches bias. One person scoring vendors in isolation will always tilt toward their own priorities. The verification step, where engineering, management, operations, and quality all review and challenge scores, creates accountability. If management wants to override engineering’s low score on requirements compliance, they have to do it on the record, in the matrix, with their name on it.
That accountability changes behavior. Not every time. But often enough that it’s worth the extra half-day the verification meeting takes.
Where Does Kepner-Tregoe Connect to GAMP Validation and Regulatory Requirements?
FDA 21 CFR 820.50 requires medical device manufacturers to evaluate and select suppliers based on their ability to meet specified requirements. That’s not a suggestion. It’s a regulatory obligation. The evaluation has to be documented.
KT scoring gives you a documented, auditable vendor evaluation that directly maps to this requirement. Every MUST is traceable to a regulatory or quality requirement. Every WANT score is recorded with its weight and rationale. The verification step creates a cross-functional record of the decision-making process.
For GAMP-validated automation systems, this matters even more. The vendor’s ability to produce compliant documentation, execute proper FAT/SAT protocols, support IQ/OQ/PQ activities, and manage changes through a controlled process isn’t a nice-to-have. It’s embedded in your validation lifecycle. Picking a vendor who can’t deliver on those items means your validation strategy is built on a foundation that won’t hold.
ISO 13485 Clause 7.4.1 reinforces this by requiring that the type and extent of supplier controls are based on the supplier’s impact on product quality. Automation equipment for medical diagnostic manufacturing sits at the high end of that scale. The controls should match.
When Does Kepner-Tregoe Vendor Scoring Add the Highest Value?
KT scoring is worth the effort on equipment procurements where the wrong choice creates regulatory, financial, or timeline damage that’s difficult to reverse. Not every purchase needs this level of rigor. Buying a standard off-the-shelf conveyor? Standard procurement works fine.
But for these scenarios, skipping structured scoring is a real risk.
- Custom automation systems for IVD manufacturing lines where GAMP documentation is required
- Capital equipment purchases above $500K where change orders could represent 20-40% of the original contract value
- Any project where multiple stakeholders (engineering, operations, management, quality) have competing priorities on vendor selection
- Vendor selections in medical diagnostic manufacturing where 21 CFR 820.50 supplier evaluation documentation will be reviewed during FDA inspections
- Projects with tight timelines where a vendor delay directly impacts product launch or production capacity
If your automation project involves consumable diagnostics, lab components, or any Class I/II/III device where validation compliance isn’t optional, KT scoring isn’t overhead. It’s insurance that costs a fraction of what a failed vendor selection costs.
The Scoring Method Costs a Day. The Wrong Vendor Costs a Year.
Running a proper Kepner-Tregoe analysis on automation vendors takes about a day of focused cross-functional work. Defining MUSTs. Weighting WANTs. Scoring vendors. Verifying across departments. Maybe a half-day follow-up if stakeholders can’t align on weights.
Choosing the wrong vendor because you skipped that day? That costs 12 to 18 months of schedule delays, six-figure change order totals, validation rework, production revenue loss, and internal engineering hours that should have been spent on the next project.
Common Questions About Kepner-Tregoe Vendor Scoring
How is Kepner-Tregoe different from a standard vendor comparison spreadsheet?
Night and day. A standard spreadsheet lines up prices and maybe a few feature checkboxes. KT separates mandatory qualifications (MUSTs) that eliminate vendors from weighted preferences (WANTs) that rank them. The verification step, where multiple departments review and challenge scores, is what makes it an actual decision framework instead of a dressed-up price list.
Can a vendor with the lowest price still win under KT scoring?
Absolutely, if they pass every MUST and score competitively on the weighted WANTs. KT doesn’t penalize low prices. It prevents low prices from overriding everything else. A well-priced vendor who also meets quality, documentation, and capability requirements will score well. The vendors who lose are the ones hiding capability gaps behind an attractive number.
How long does a proper KT vendor scoring process take?
Plan for a day, sometimes two. Half a day to define MUSTs and weight WANTs with your cross-functional team. Half a day to score vendors against the matrix. Then a verification meeting where stakeholders challenge and adjust scores. For a $1M+ automation project, that’s less than 0.1% of the project timeline spent preventing the single decision with the highest failure impact.
Is Kepner-Tregoe required by FDA regulations?
The framework itself isn’t mandated. But 21 CFR 820.50 requires documented supplier evaluation based on ability to meet specified requirements. KT gives you exactly that, a structured, documented, auditable evaluation process. During an FDA inspection, pulling out a completed KT matrix is a clean answer to “how did you select this vendor?” A purchase order and three price quotes is not.
What happens if management and engineering disagree on vendor scores?
That’s actually the point. KT surfaces disagreement before the purchase order, not after the vendor fails. Run the matrix with different weight scenarios. Show management what happens when cost dominates versus when technical capability dominates. I’ve been in rooms where this conversation saved projects. I’ve also been in rooms where management overrode engineering anyway, and the project came in 18 months late. The matrix doesn’t force a decision. It forces transparency about the tradeoffs.
The Bottom Line
The lowest bid almost always costs more. The scoring method tells you why before it’s too late.
If you’re evaluating automation vendors for a medical diagnostic manufacturing project and want a structured approach to getting the selection right the first time, a Manufacturing Automation Assessment is the logical starting point. It maps your project requirements to the scoring criteria that actually predict vendor success.