Subtract to Ship is the discipline of removing work that does not earn its place. Applied to MedTech, it means running the startup on a small number of load-bearing activities — the ones that move the device toward a real market and a clean certification — and refusing everything else. The MedTech version has four passes: Purpose, Design Freeze, Evidence, and Operations. The discipline is week-to-week, not a one-time exercise. Done honestly, it gives a small team the focus to ship a medical device on a startup budget without cutting anything required by the Regulation.

By Felix Lenhard and Tibor Zechmeister. Last updated 10 April 2026.

TL;DR

  • Subtract to Ship is an operating methodology, not a regulatory framework. It started with startups that had nothing to do with medical devices and transfers into MedTech because the discipline of removing work that does not earn its place is domain-independent.
  • MedTech inflates work in predictable ways — feature bloat, parallel workstreams, premature scaling, template-driven documentation, and meetings that substitute for decisions.
  • The four passes applied to MedTech operations are Purpose, Design Freeze, Evidence, and Operations. Each one subtracts a different category of drag.
  • The design freeze is the single highest-leverage move in a MedTech startup. Every week without a freeze multiplies downstream cost.
  • Subtract to Ship is enforced week-to-week on the team calendar, not once at the kickoff. A weekly subtraction audit keeps the company lean.
  • The methodology is compatible with every obligation in Regulation (EU) 2017/745. It cuts waste. It does not cut compliance and it does not cut safety.

A pandemic story that predates any medical device

Years before Felix ever worked on a MedTech project, he was inside a consumer brand called Vulpine when the pandemic hit. The company made premium cycling apparel. On paper it had dozens of product lines, a large website, an active social presence, a production pipeline, a community, and the normal overhead of a brand that had been building for years. Most of that was drag, and nobody noticed until the drag became fatal.

When the pandemic shut down the economy overnight, Vulpine could not keep running the way it had been. The team stripped the company to the few things that could carry revenue through the next quarter. Whole product categories came out. Marketing channels were cut. Meetings were cancelled. The website was reduced. What stayed was the small number of activities that were actually keeping the lights on. Three things became obvious once the stripping was done. Nobody had been using most of the side features. Scarcity forced sharper decisions than abundance ever had. And the team's stubbornness compounded once the scope was small enough to actually finish.

That is where Subtract to Ship started as a named discipline. Not as a framework, but as a survival move that worked. Felix spent the next several years applying the same move to startups across industries — forty-four of them so far, mostly outside regulated domains — and the pattern held every time. Removing work that did not earn its place was cheaper, faster, and more honest than adding more work to compensate for the work that was already drag.

MedTech looks nothing like a consumer apparel brand on the surface. Underneath, the disease is identical. Startups in MedTech inflate work for the same reasons startups in every other domain do — the work looks like progress, the work feels like diligence, the work fills the week, and nobody on the team is incentivised to stop it. The only difference is that a MedTech startup that inflates work runs out of money faster, because the floor cost of the regulatory project does not leave any room for waste on top of it.

This post is the operating methodology side of Subtract to Ship in MedTech. The regulatory angle — how the same discipline applies to the certification project itself — is covered separately in the Subtract to Ship framework for MDR. This one is about how you run the company week to week.

What Subtract to Ship actually is

Subtract to Ship is a methodology for shipping under constraints by removing work that does not earn its place, instead of adding work that looks productive. That is the whole definition. Everything else is how the definition is applied.

Three things flow from it.

First, every activity in the company must be defensible. If you cannot explain why a task is on the board this week — what it moves, what it unblocks, what it kills — the task is drag and comes off the board. "It felt important" is not a defence. "A consultant recommended it" is not a defence. "We have always done it" is not a defence.

Second, subtraction is continuous, not one-off. A kickoff meeting where you cut a bunch of items is useful but temporary. The next week's meetings will re-add everything you cut unless subtraction is built into the rhythm of the company. The discipline lives in a recurring audit, not in a single moment.

Third, subtraction is not the same as cutting corners. Cutting corners means removing work that was required. Subtraction means removing work that was never required in the first place. The test — always — is whether the activity traces back to something real: a Regulation obligation, a clinical commitment, a contractual requirement, a specific customer promise, or a financial constraint that cannot be negotiated. If it traces to one of those, it stays. If it does not, it goes.

That discipline is industry-independent, which is why it transferred from a consumer brand in a pandemic into a MedTech startup without any translation layer.

Why MedTech inflates work more than most industries

MedTech has structural forces that push startups toward inflation. Naming them is the first step to resisting them.

The fear premium. Regulatory work feels dangerous because it is dangerous, and the founder instinct when something is dangerous is to do more of it. More documents. More review cycles. More cross-checks. The problem is that fear-driven work is rarely targeted at the real risk — it is targeted at the anxiety. Ten extra pages in a procedure nobody follows do not reduce regulatory risk. They increase it, because an auditor can see the gap between the document and the practice in the first ten minutes.

The template trap. Because the field is regulated, template packs exist. Hundreds of documents, pre-filled, ready to be renamed. They look like a shortcut and become a trap — the team ends up owning a QMS that describes a company that does not exist, and every audit finding is a finding against a process the company never actually ran.

The consultant load-out. A consultant who is paid by the deliverable has an incentive to recommend more deliverables. Not out of malice — out of arithmetic. A lean project is a smaller invoice. A founder who does not push back gets the full menu.

The parallel workstream problem. MedTech teams try to run everything at once — hardware, software, QMS, clinical, regulatory, fundraising, customer development, team building. Six parallel workstreams with one founder result in six half-finished workstreams and zero ships. The fix is sequencing, not more heroics.

Meetings that replace decisions. Because regulatory work is ambiguous, teams drift into meetings to "align" instead of to decide. Three meetings a week, no closed questions, no actions with owners. The meetings feel productive and produce nothing. This is the single most common form of MedTech drag Felix sees across startups.

Subtract to Ship is the discipline of naming these patterns out loud and refusing to be pulled into them.

The four passes applied to MedTech operations

The methodology runs through four passes. The regulatory version of the passes is covered in the companion post — Purpose, Classification, Evidence, Operations — with the Regulation as the referee. The operations version sits alongside it and addresses how the company runs. The two versions reinforce each other; this post covers the operations side.

Pass 1 — The Purpose Pass (operations version)

The first pass asks: what is the startup actually trying to build, for whom, and what is the one number that says it is working?

Every MedTech startup that inflates work has this pass missing. The founder can list ten things the device might be useful for, three market segments it might enter, and two clinical specialties it could apply to. The list is a signal that the purpose is not set. Until it is, every week will produce work on all of them, because nothing has been cut.

The operations output of the Purpose Pass is a short paragraph: one buyer, one clinical setting, one primary benefit, and the single metric that says the device is working for that buyer in that setting — time saved per case, patients processed per day, complications avoided per month. Everything that is not on that sheet is drag.

This is the operations cousin of the regulatory Purpose Pass. The regulatory version writes the intended purpose under MDR Article 2(12). The operations version writes the business purpose. The two should agree. If the business purpose names a buyer the intended purpose cannot serve, one of them is wrong and the company is paying for the disagreement.

Pass 2 — The Design Freeze move

The single highest-leverage move in a MedTech startup is the design freeze. This is the pass most first-time founders resist the longest and regret the most.

The design freeze is the moment you stop changing the device. Not permanently — but for long enough that the rest of the project can be built against a fixed target. Risk management under EN ISO 14971:2019+A11:2021 needs a stable design to identify and control risk. Technical documentation needs a stable design to describe. Clinical evaluation under Article 61 needs a stable design to evaluate. Verification and validation testing needs a stable design to test. Every week the design moves, all of that work moves with it — and most of it has to be redone.

Felix's rule: the design freeze is the most expensive week in the company if you delay it, and the cheapest week in the company if you call it. Founders delay it because they are still chasing improvements. Every improvement feels like it is making the product better. It is, marginally — and in exchange, it is setting every downstream workstream on fire. The right move is to freeze on a design that is good enough to ship, then ship, then improve in the next version under change control. The wrong move is to keep adjusting because each individual adjustment is cheap, ignoring the fact that the cumulative cost compounds.

Subtraction in this pass is brutal. Every feature still under active debate comes out unless it is load-bearing for the primary benefit. Every optional mode gets cut. Every configurability that was added "in case a customer wants it" gets removed until a customer actually asks for it in writing. The device that emerges from the Design Freeze is smaller, sharper, and ships.

Pass 3 — The Evidence Pass (operations version)

The third pass asks: what is the minimum evidence the company needs to make its next decision, and what is the cheapest path to that evidence?

Startups confuse evidence gathering with activity. They run surveys that will not change the plan, build dashboards that nobody reads, and commission research reports that arrive too late to matter. The operations Evidence Pass forces a different question: what is the specific decision this evidence is meant to inform, and when does that decision have to be made?

If the decision is "do we pursue this buyer segment," the evidence you need is three to five honest buyer conversations, not a market research report. If the decision is "is the workflow realistic," the evidence is one day of shadowing in a real ward, not a persona document. If the decision is "can this clinical claim be defended," the evidence is a literature review and one conversation with a clinical partner, not a standalone study.

Subtraction in this pass means cutting every evidence-gathering activity that is not wired to a specific decision on a specific date. "We want to know more" is not a reason to run work. "We need this answer by the end of next week to decide X" is.

This operations version sits alongside the regulatory Evidence Pass that decides the cheapest legitimate pathway to clinical evidence under Article 61. Both passes ask the same question — what is the minimum evidence needed — in different registers.

Pass 4 — The Operations Pass

The fourth pass is about how the company actually runs its week. Headcount, meetings, tools, processes, reporting, calendar.

The test: for every recurring activity in the company — every weekly meeting, every tool subscription, every reporting cadence, every role — what would break if it stopped? If the honest answer is "nothing," the activity is drag and comes out. If the honest answer is "the regulatory file would be behind" or "the clinical partner would lose trust" or "the QMS training record would be incomplete," the activity is load-bearing and stays.

Subtraction in this pass means running the company on the smallest possible surface area. Small team — one regulatory lead, one QMS owner, one clinical owner, one PRRC under Article 15, and engineering built for the specific device. Small tool stack — pick the minimum number of systems the team actually uses, cancel the rest. Small meeting load — every meeting has a specific decision to make, or it does not run. Small reporting cadence — the investor update is monthly and short, the team standup is weekly and short, everything else is on demand.

EN ISO 13485:2016+A11:2021 asks for a QMS proportionate to the organisation and the risk class. MDR Article 10(9) uses the word "proportionate" directly. The Operations Pass is what turns "proportionate" from a word into a lived weekly rhythm.

The discipline week-to-week

Subtract to Ship fails when it is treated as a one-off. A single subtraction exercise at the kickoff meeting wears off in three weeks. The work comes back. The meetings come back. The parallel streams come back. Unless subtraction is built into the recurring rhythm of the company, entropy wins.

The rhythm Felix runs with founders is simple.

Weekly — the subtraction audit. Fifteen minutes on a Friday. Walk through everything the company did that week. For each item, ask whether it earned its place — whether it moved the primary metric, unblocked a load-bearing decision, or satisfied a required obligation. Anything that did not earn its place gets named out loud. Not punished — named. Next week the team notices the pattern earlier.

Weekly — the week-ahead cut. Ten minutes on a Monday. Look at the planned work for the week. Cut at least one item. Even if the week looks lean, cut one item. The discipline is not about the specific item — it is about forcing the team to defend the week against the question "why this?"

Monthly — the tool and meeting audit. Thirty minutes. Walk through every recurring meeting and every tool subscription. Ask the honest question: what breaks if this stops? Cancel the ones where nothing breaks.

Quarterly — the scope review. Two hours. Walk through the purpose, the design freeze status, the evidence decisions in flight, and the operations load. Re-subtract. What was load-bearing last quarter may be drag this quarter, because the company has moved.

These rituals are cheap. They produce disproportionate results because they prevent inflation rather than cleaning it up after the fact.

Common founder mistakes

Mistake 1 — treating Subtract to Ship as permission to skip required work. The Regulation does not care about your operating methodology. If MDR Article 83 requires a PMS system, you have one. If Article 15 requires a PRRC, you have one. Subtract to Ship cuts drag, not obligation. The founders who try to subtract their way out of compliance fail fastest.

Mistake 2 — subtracting the wrong things because they are the easy things. Cancelling the coffee budget is not subtraction. Cancelling the weekly customer conversation is. Subtraction is hard precisely because the things that most need cutting are the things that feel productive — the extra documentation, the extra meeting, the extra review cycle. The comfortable cut is usually the wrong cut.

Mistake 3 — one-off subtraction. Treating the kickoff cut as the whole methodology. It is not. Without the weekly rhythm, the work comes back and the founder wonders why the startup is inflated again six weeks later.

Mistake 4 — subtracting the design freeze. The founder refuses to freeze because "one more improvement" always feels worth it. Every week of delay compounds downstream. The honest fix is to call the freeze early, ship the version that is good enough, and put the improvements into the next release.

Mistake 5 — confusing small team with no team. Subtract to Ship does not mean one founder doing everything. It means a small team of load-bearing roles with no drag hires. The difference is consequential. A missing role is not subtraction. It is a gap.

The Subtract to Ship angle (explicit framework summary)

Because this post is the operations angle, here is the framework in one compact form so a reader who lands only on this page gets the whole thing.

  • Definition. Subtract to Ship is the discipline of shipping under constraints by removing work that does not earn its place, instead of adding work that looks productive.
  • Origin. Developed by Felix Lenhard across 44 startups, starting from a pandemic-era stripping exercise at the Vulpine brand and refined in every startup since. Domain-independent by design.
  • Core test. For every activity: does it trace to a specific obligation, a load-bearing decision, a primary metric, or a contractual commitment? If yes, keep it. If no, cut it.
  • The four passes, operations version. Purpose (what are we actually building and for whom), Design Freeze (when does the device stop moving), Evidence (what is the minimum to decide), Operations (what runs the company week to week).
  • The rhythm. Weekly subtraction audit, weekly week-ahead cut, monthly tool and meeting audit, quarterly scope review.
  • The red lines. Compliance stays. Safety stays. Contractual commitments stay. Subtraction only touches drag.
  • The output. A small, focused team running a small number of load-bearing activities on a fixed design toward a verified market. Slow is smooth. Smooth is fast.

The regulatory companion post — the Subtract to Ship framework for MDR — runs the same discipline against the certification project itself.

Reality Check — Where do you stand?

  1. Can you name, in one paragraph, the single buyer, the single clinical setting, the single primary benefit, and the single metric your device is targeting? If there is a list, the Purpose Pass has not been run.
  2. Is your device currently under design freeze, or is the design still moving every week? If it is moving, what is the specific date it stops?
  3. For every recurring meeting on your calendar, can you name the decision it makes? If a meeting has no decision, why is it still on the calendar?
  4. For every tool subscription your team pays for, what would break if it was cancelled tomorrow?
  5. Do you run a weekly subtraction audit, or did you cut work once at kickoff and assume it would stay cut?
  6. Is every activity in your current week defensible by tracing back to a load-bearing obligation, decision, or metric?
  7. When you imagine cutting one item from next week's plan, which item is the first to go — and why is it still on the plan this week?

Frequently Asked Questions

Is Subtract to Ship a MedTech methodology? No. It is a general operating methodology developed by Felix Lenhard across 44 startups, most of them outside regulated industries, starting from a pandemic-era stripping exercise at a consumer brand. This post applies it specifically to MedTech operations. The regulatory application is covered in the companion post on the Subtract to Ship framework for MDR.

How is this different from lean startup? Lean startup is a build-measure-learn loop with hypotheses tested against customers. Subtract to Ship is narrower and more specific — it is the discipline of removing work that does not earn its place. The two are compatible. Subtract to Ship sits inside any operating system a MedTech startup chooses, and forces the discipline of cutting drag on a recurring basis.

Does Subtract to Ship conflict with the Regulation? No. Every obligation in Regulation (EU) 2017/745 and every requirement in the referenced harmonised standards stays in scope. The methodology cuts work that was never required in the first place — template bloat, parallel workstreams, fear-driven documentation, meetings without decisions. The Regulation is the referee that decides what cannot be cut.

When is the right time to call a design freeze? Sooner than the founder wants to. The test is whether the current design is good enough to ship the primary benefit for the primary buyer. If it is, freeze. Remaining improvements go into the next release under change control. The cost of delay compounds across every downstream workstream — risk management, technical file, clinical evaluation, verification and validation — so the founder who freezes early recovers weeks of runway the founder who delays never gets back.

How small can a MedTech team actually be? For a Class I or Class IIa device run with discipline, the load-bearing roles are a regulatory lead, a QMS owner, a clinical owner, a PRRC under Article 15 (which can be contracted externally for micro and small enterprises), and engineering built for the specific device. Everything else is bought on demand. The discipline is not "hire nobody." The discipline is "hire only load-bearing roles."

Sources

  1. Regulation (EU) 2017/745 of the European Parliament and of the Council of 5 April 2017 on medical devices, consolidated text. Articles referenced: Article 10(9) on proportionate QMS; Article 15 on the Person Responsible for Regulatory Compliance; Article 61 on clinical evaluation; Article 83 on post-market surveillance. Official Journal L 117, 5.5.2017.
  2. EN ISO 13485:2016 + A11:2021 — Medical devices — Quality management systems — Requirements for regulatory purposes.
  3. EN ISO 14971:2019 + A11:2021 — Medical devices — Application of risk management to medical devices.

This post is part of the MedTech Startup Strategy & PMF series in the Subtract to Ship: MDR blog. Authored by Felix Lenhard and Tibor Zechmeister. The operations discipline described here is the week-to-week rhythm that makes the regulatory project possible on a startup budget. Slow is smooth. Smooth is fast.