A founder I know ran a thorough qualification. ISO 13485, audit history, capability range, three reference calls. Scored every supplier against the same rubric. Picked the top scorer.

Six months later the design changed. The supplier quoted a 14-week lead time for a geometry revision, then stopped responding after the second follow-up. The qualification had been thorough. It had also missed the thing that mattered.

Supplier selection has two layers. Most founders only see one. #

The first is the technical layer: can they make the part? Certifications, capability range, MOQ, audit history, sample quality. You can verify all of it through a packet. This is what most RFQs are built to assess. It’s necessary, but it’s not the whole picture.

The second is the adaptive layer: can we work together as things change? There’s no spec sheet for this. You can’t verify it through credentials. You build the understanding through interaction, and that interaction has to start during evaluation, not after the PO is signed.

Ronald Heifetz at Harvard Kennedy School draws a line between technical challenges — problems with known solutions where you find someone who has the answer and buy it — and adaptive challenges, where the solution has to be built through learning and relationship. Picking a supplier has both. The mistake is treating the whole thing as technical.

The technical layer: what you can buy #

ISO 13485 or equivalent. Device class-specific experience. Traceability practices. Audit trail. For cross-border sourcing, any relevant FDA registration.

Run these checks. But two suppliers who pass the same technical checks can behave completely differently when something unexpected happens. The technical layer tells you what they can do today. It says nothing about what they’ll do when today’s plan breaks.

The adaptive layer: five things to read during evaluation #

You can’t score this on a rubric. But you can read it during evaluation if you know where to look.

First: how they handle gaps in the RFQ. Most RFQs have open questions. A supplier who asks good ones during quoting will flag problems early during production. One who quotes everything without questions is either drawing on deep experience or guessing. Worth asking which.

Second: who you actually talk to. Owner or senior engineer, or only a sales contact? At pre-Series A, owner access matters more than a polished sales process. You will need to call someone at 9pm when something is wrong. Make sure that person answers.

Third: how they describe past design changes. Ask directly: what’s the most disruptive change a customer threw at them, and what happened? A supplier who names a specific example, with detail, has been in those situations. One who gives you a vague answer probably hasn’t.

Fourth: how they respond to a request for a production flow diagram. Not a capability statement — the actual step-by-step process, including any outsourced steps. A supplier who shares it without hesitation is one who expects to be seen. That posture carries through to production.

Fifth: how they handle an awkward question. Ask about their largest quality escape in the last two years. Watch the response time. Watch whether the answer is specific or a deflection. If they’re evasive during evaluation, they’ll be evasive when something breaks.

The procurement frame vs. the hire frame #

Most founders approach supplier selection like procurement: send the RFQ, score the responses, pick the best. That frame works when you know exactly what you need and nothing will change.

Pre-Series A medtech is the opposite. Your design will change. Volume will shift by orders of magnitude. The supplier who scores highest on today’s spec may be useless when the spec changes next quarter.

The better frame is a hire. You’re bringing on a partner for an unknown future, not buying a commodity. You wouldn’t hire a senior engineer off a resume score alone — you’d talk to them, see how they think, ask how they’ve handled hard situations. The RFQ is the resume. Most founders never get to the interview.

Buy the technical. Build the adaptive.

What to do with your current shortlist #

Go back to each serious candidate and ask the five questions above. In a conversation, not a formal audit. Watch how they respond. The answers matter less than the posture.

If you want to work through a specific shortlist, book a 1:1. Most founders who book come with two or three candidates and leave with a decision.

This article appears in the Agile Supplier Field Guide — a strategic guide for pre-Series A medtech founders.