UDI compliance under MDR Article 27 and Annex VI Part C has two cost shapes: fixed one-time costs to set the system up, and recurring costs to keep it running. The fixed side covers the issuing entity joining fee, the first label redesign, the ERP and labelling software integration, the first Basic UDI-DI and UDI-DI allocation, and the first EUDAMED data submission under Commission Implementing Regulation (EU) 2021/2078. The recurring side covers annual issuing entity membership fees, label and packaging revisions when the product changes, EUDAMED data maintenance, and the engineering time for every new UDI-DI triggered by a variant, a packaging change, or a software release that touches identification. The dominant cost for almost every startup is team time, not vendor invoices. A first-device UDI rollout done cleanly is a few weeks of focused work across regulatory, engineering, operations, and label design. Done badly it becomes a six-month drag with scrapped label stock at the end.

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

TL;DR

  • UDI costs split cleanly into fixed one-time costs and recurring costs. Both have to be in the budget from day one.
  • Issuing entity fees exist at all four designated entities — GS1, HIBCC, ICCBBA, IFA — and recur annually. The specific numbers change; confirm with each entity directly before signing off a budget.
  • Label and packaging changes are the single largest cash cost at launch for most startups. Scrapped stock from a non-compliant first print run is the common failure mode.
  • EUDAMED data entry is not expensive in cash, but it is expensive in hours when done for the first time. The Annex VI Part B fields are detailed and require cross-functional input.
  • Ongoing maintenance is a permanent line item, not a one-off. Every variant, every packaging revision, every software release that touches identification can trigger a new UDI-DI.
  • The dominant cost at almost every startup is team time. A disciplined rollout costs a few weeks of focused cross-functional work. A chaotic one costs months and scrapped material.

A founder opens a budget spreadsheet and cannot find the UDI line

The scenario repeats across first-time MedTech startups. The founder has costed the Notified Body fees, the QMS software, the clinical evaluation, the biocompatibility testing, and the risk management workshops. Then someone on the team asks what the UDI line is. The founder looks at the spreadsheet and realises UDI has been assumed to be free — a box on a label, a row in a database, nothing that should need its own budget. A few weeks later, the first real conversation with the print vendor, the ERP integrator, and the issuing entity produces numbers that were nowhere in the plan. The label vendor quotes a redesign. The integrator quotes work on the labelling software. The entity quotes an annual fee. The internal team time adds up faster than any of those.

This post fixes that gap. It walks through the cost categories honestly, at general framing rather than specific numbers that would go stale, and it separates the cash costs from the team-time costs that almost always dominate. The goal is a founder who closes this tab with a UDI line on the budget that holds up when real invoices start arriving. For the conceptual orientation to UDI, read what is UDI?. For the entity decision that drives most of the cost shape, read how to obtain a UDI by choosing an issuing entity. For the carrier side, read UDI carrier requirements.

The cost categories

Before the numbers, the categories. A realistic UDI budget has seven lines, and every one of them shows up on every project. Treating any of them as zero is how a UDI plan goes over budget.

  1. Issuing entity fees. Joining fee plus annual membership with the chosen entity under MDR Article 27(2).
  2. Label and packaging changes. Design, proofing, and print-run costs for labels and packaging that carry the UDI in both AIDC and HRI forms at every required packaging level.
  3. ERP and labelling software integration. Internal plumbing so the UDI-DIs, UDI-PIs, and Basic UDI-DIs flow from the master data system into the labels, the shipping documents, and the EUDAMED submission.
  4. EUDAMED data entry effort. First-time preparation and submission of the Annex VI Part B data elements under Commission Implementing Regulation (EU) 2021/2078.
  5. Direct part marking engineering (only where applicable). Design, validation, and production-line integration for laser marking or equivalent on reusable devices.
  6. Ongoing maintenance. Annual entity fees, periodic data updates, new UDI-DIs for variants and packaging revisions, and SaMD re-identification for software releases.
  7. Team time. The cross-functional hours from regulatory affairs, engineering, operations, label design, and QA to run all of the above.

Every founder's instinct is to look at lines 1 through 6 first because they map to invoices. Line 7 is the one that decides whether the project is a few weeks or several months, and it is the one that founders systematically under-budget.

Issuing entity fees at general framing

The four Commission-designated issuing entities — GS1, HIBCC, ICCBBA, IFA — each charge for the right to allocate identifiers under their system. The specific numbers change periodically, and each entity should be contacted directly to confirm the current schedule before a budget is signed off. The structural shape of the fees is more stable and more useful for a planning conversation.

GS1 is organised through national member organisations and fees are set by each national organisation. Most charge a one-time joining fee plus an annual membership fee that scales with company revenue. The smallest startups typically fall into the lowest revenue tier. Companies growing past the smallest tier move up the revenue bands over time. For most European startups this is a modest line in year one that grows with the company.

HIBCC charges a flat annual licensing fee for a Labeler Identification Code, independent of company revenue. Flat fees favour larger companies at the crossover point. Most startups are well below that crossover.

ICCBBA charges a licensing and registration fee structure for the ISBT 128 system. For devices in ICCBBA scope — products of human origin and directly coupled devices — the fee is rarely the deciding factor.

IFA charges per registered product through the German pharmacy data catalogue. Wide portfolios multiply the fee; narrow portfolios minimise it.

Two structural observations matter more than any specific number. First, the entity fees are an operating cost, not a capital cost — they recur every year for as long as the device is on the market. Second, the fee difference between entities is, for almost every startup, much smaller than the operational difference between entities. Choosing a cheaper entity whose carriers the European distribution channel cannot read is an expensive mistake dressed up as a cheap decision. The entity selection logic is walked through in the companion post on obtaining a UDI by choosing an issuing entity.

Label and packaging changes

The label line is usually the largest first-year cash cost after the entity fee, and the one most prone to unpleasant surprises. A compliant UDI label carries the UDI in both AIDC and HRI forms at the same time, at every packaging level up to but not including the shipping container, under Annex VI Part C.

There are three cost components inside the label line.

Design and proofing. A label designer who has shipped MDR-compliant labels before will build the AIDC and HRI together, with the correct Application Identifier syntax (for GS1), a legible HRI block, and the right content at each packaging level. A designer who has not done this before will produce something that looks fine and fails on a technicality — compressed HRI, missing packaging level, wrong data identifier. The premium for a designer who has done it is small and recovers itself on the first audit review.

Proof and validation. Every carrier design has to be proofed by scanning with the same equipment the downstream supply chain will use. A print proof that reads perfectly in the design studio and fails at a hospital scanner is a common failure mode. Proofing on real target hardware is a small line item that prevents a large scrap line.

Print runs. The physical print run is where scrap appears if the design was wrong. A founder who orders a full-year stock of labels against a non-compliant design and discovers the error at Notified Body review scraps the stock. The mitigation is to run a small first batch, validate it end-to-end, and only then commit to volume. The Subtract to Ship instinct applies: the smallest print run that validates the design is the right first print run.

Packaging follows the same logic. Every packaging level gets its own UDI-DI and its own label, and every level has to be designed, proofed, and printed. The larger the packaging hierarchy, the larger the label line.

EUDAMED data entry effort

The EUDAMED submission under Commission Implementing Regulation (EU) 2021/2078 is cheap in cash and expensive in hours the first time. There is no submission fee. There is a considerable data preparation cost.

The Annex VI Part B fields are detailed — device identification, risk class, manufacturer SRN, authorised representative where applicable, Notified Body where relevant, substances of concern, IFU URLs, clinical investigation data, and a long tail of fields that each require a source inside the organisation to provide an accurate answer. Assembling these for the first device means a regulatory person walks through the list, identifies which person owns each field, collects the answer, verifies it, and enters it. The walk-through is the cost.

Once the first device is in, subsequent devices inside the same family reuse most of the groundwork. The organisational structure is set. The SRN is live. The authorised representative fields are populated. The first submission is the expensive one; the second and third are materially cheaper. A startup that plans only for the first submission underestimates nothing. A startup that assumes the first submission will cost the same as the fifth overestimates both and underprepares for the gap between them.

Ongoing maintenance

UDI is not a launch project that finishes at first placement on the market. It is a permanent operational process that has to be resourced as long as the device is on the market.

Annual entity fees. The membership or licensing fee at the chosen issuing entity recurs every year, for every year the device is on the market.

Data updates. MDR Article 29 requires the manufacturer to keep the UDI database information up to date. Changes to the manufacturer address, to the authorised representative, to the Notified Body certificate, to the IFU URL, to substances of concern declarations — all of these have to be pushed into EUDAMED in a timely way. The effort per update is small; the number of updates per year across a growing portfolio is not.

New UDI-DIs for variants and packaging. Every new variant, every new packaging configuration, and every meaningful design change that affects identification requires a new UDI-DI, a new label, and a new Annex VI Part B submission. A company that ships one device becomes a company that ships several, and the UDI line grows with the portfolio.

SaMD re-identification. For software that is itself a medical device, Annex VI Part C triggers a new UDI-DI on any modification that changes original performance, safety, intended use, interpretation of data, or other identification-related aspects. Minor revisions share the UDI-DI with a new UDI-PI. The pace of software releases determines the UDI-PI cadence, and the content of each release determines whether a new UDI-DI is required. Companies shipping SaMD on a continuous delivery cadence need this process automated or it becomes a perpetual manual burden. The companion post on UDI for software medical devices goes into the SaMD specifics.

Direct part marking line maintenance (where applicable). Reusable devices with direct part marking need the marking process kept in spec as production lines change. Laser parameters drift, marking positions move, and validation has to follow.

Treating these as one-time launch costs is the common budgeting error. They are permanent.

The team-time cost

The single largest cost on almost every UDI project is team time, and it is the one founders almost always under-budget.

A first-device UDI rollout done cleanly is a few weeks of focused cross-functional work. That means a regulatory person running the entity decision, the Basic UDI-DI allocation, and the EUDAMED data preparation. An engineering person owning the UDI-PI composition, the data carrier choice, and any direct part marking. An operations person integrating the identifiers into the ERP, the labelling software, and the production workflow. A label design partner producing compliant AIDC and HRI carriers at every packaging level. A QA person folding the whole thing into the technical documentation.

These four or five people are not full-time for the duration, but they each need to be available at specific handover points, and the handovers have to line up. A project that serialises the work — entity choice waits for label design waits for ERP integration waits for EUDAMED submission — takes months. A project that parallels the work with clear handover points takes weeks.

The other team-time cost is the rework when any step is wrong. A mis-allocated Basic UDI-DI that has to be revised after the technical file is already written regenerates several documents. A label print run against a non-compliant design regenerates the label stock and everything printed on top of it. A missing Annex VI Part B field at submission time blocks the device registration and pulls the team back to fix it while every other workstream waits. Rework is the hidden cost that turns a few weeks into a few months.

The implication for budgeting is that the team-time line on a UDI project should be the biggest line by hours and should include a buffer for the rework that almost always happens on a first device. Pretending the rework will not happen is how projects slip.

Common mistakes

  • Budgeting only the cash items. Entity fees and label prints are visible. The dominant cost — team time — is invisible on most budgets and produces the biggest overruns.
  • Treating the entity fee as a one-off. The fee recurs every year. The line belongs in the operating budget, not only the launch budget.
  • Ordering a full-year label print before validating the design end-to-end. The scrap cost when the design fails review is the single most common UDI surprise.
  • Assuming EUDAMED is expensive because it is regulatory. Cash-wise it is not. Hour-wise for the first submission it very much is.
  • Forgetting direct part marking when the device is reusable. Engineering and validation cost that surfaces late is the most disruptive kind.
  • Treating software UDI as a label problem. SaMD re-identification is continuous. Without automation it becomes a permanent drag on every release.
  • Designing labels with a vendor who has never shipped MDR. The premium for an experienced vendor is smaller than the cost of one rework cycle.
  • Skipping the internal handover map. A project that has all the right people but no clear sequence of handovers runs three times as long as one that has the same people with the sequence written down.

The Subtract to Ship angle on UDI cost

UDI cost is a clean case for the Subtract to Ship framework for MDR because the work that survives is finite and traceable, and the waste is the work that does not trace.

The real work is: pay the entity fee for the chosen entity, design compliant labels at every required packaging level with both AIDC and HRI, integrate the identifiers into the ERP and labelling software, submit the Annex VI Part B fields to EUDAMED under (EU) 2021/2078, engineer direct part marking where the device is reusable, maintain the data and the identifiers as the portfolio and the product evolve, and budget enough team time for the handovers and the rework. That list is the whole list.

Everything else is a candidate for removal. The dual-entity parallel rollout for a first device. The custom internal numbering scheme layered on top of the issuing entity codes. The speculative RFID deployment for a supply chain with no RFID readers. The label redesign project that optimises typography while missing the HRI legibility test. The EUDAMED consultant hired to do work the internal team has already done. The "UDI readiness workshop" that does not produce a single allocated identifier. None of these trace back to Article 27, Annex VI Part C, or (EU) 2021/2078. They come out of the budget.

Reality Check — Is your UDI budget honest?

  1. Does your budget include an explicit line for the annual issuing entity fee, recurring every year the device is on the market?
  2. Does the label budget include a validation print run before the full production print run?
  3. Have you confirmed the current fee schedule with the chosen entity directly, not from a third-party article?
  4. Does the budget include team time from regulatory, engineering, operations, label design, and QA, costed as hours rather than assumed as free?
  5. Have you budgeted for first-submission EUDAMED hours separately from steady-state data maintenance hours?
  6. If your device is reusable, have you budgeted direct part marking engineering and validation as a separate line?
  7. If your device is SaMD, have you budgeted ongoing UDI re-identification effort into the release process?
  8. Does the budget include a rework buffer for the first-device project, or does it assume everything goes right on the first pass?

If you cannot answer six or more of these cleanly, the UDI budget is not yet realistic.

Frequently Asked Questions

What is the biggest cost line in a UDI rollout for a startup? For almost every startup, the biggest cost line is team time across regulatory, engineering, operations, and label design, not vendor invoices. A disciplined first-device rollout is a few weeks of focused cross-functional work. A chaotic one becomes months of rework. The cash costs — entity fees, label prints, ERP integration — are visible on a spreadsheet; the team-time cost is the one that decides whether the project ships on time.

Are issuing entity fees one-time or recurring? They are recurring. All four designated issuing entities — GS1, HIBCC, ICCBBA, IFA — charge annual membership or licensing fees that continue for as long as the manufacturer is allocating identifiers under their system. Treating the fee as a one-off is a common budgeting error.

How much does an EUDAMED submission cost in fees? There is no submission fee to EUDAMED itself. The cost of EUDAMED submission is the team time to assemble the Annex VI Part B data elements required under Commission Implementing Regulation (EU) 2021/2078 and to enter them correctly. The first submission is the expensive one in hours; subsequent submissions reuse most of the groundwork.

Do I need to budget for label rework on a first rollout? Yes. First-device label design on a first MDR project almost always goes through at least one revision, and an experienced budget includes the scrap cost and the team time of at least one rework cycle. Validating the design with a small print run before committing to a full production run is the cheapest way to contain the rework cost.

How does direct part marking change the budget? For reusable devices that require direct part marking under Annex VI Part C, the UDI budget includes an engineering line for the marking process — typically laser marking — plus validation that the marking survives the reprocessing cycles. This is an engineering workstream, not a labelling workstream, and it should be budgeted separately and started early.

What makes SaMD UDI costs different? Software UDI costs are continuous rather than one-off. Annex VI Part C triggers a new UDI-DI on any software modification that changes original performance, safety, intended use, or identification. Minor revisions share the UDI-DI with a new UDI-PI. For continuous-delivery software this means the UDI effort is folded into every release, and most startups that ship SaMD need the process automated rather than manual.

Is the entity fee difference a good reason to choose one entity over another? Rarely. The fee difference between entities is, for almost every startup, much smaller than the operational difference between entities. Choosing an entity whose carriers the European distribution channel cannot read to save a modest fee is an expensive decision dressed up as a cheap one. The entity choice should follow the distribution channel, not the fee schedule.

Sources

  1. Regulation (EU) 2017/745 of the European Parliament and of the Council of 5 April 2017 on medical devices, Article 27 (Unique Device Identification system, including the obligation to assign UDIs under the rules of an issuing entity designated by the Commission) and Annex VI Part C (the UDI system, including carrier requirements, packaging-level obligations, direct part marking for reusable devices, small-device exceptions, and the rules for software UDI re-identification). Official Journal L 117, 5.5.2017.
  2. Commission Implementing Regulation (EU) 2021/2078 of 26 November 2021 laying down rules for the application of Regulation (EU) 2017/745 as regards the European Database on Medical Devices (Eudamed), including the UDI database module and the Annex VI Part B data element submission. OJ L 426, 29.11.2021.
  3. Issuing entity fee schedules — GS1 national member organisations, HIBCC, ICCBBA, and IFA each publish their current fee schedules on their own websites. Current-fees verification note: the specific fee numbers at all four designated issuing entities change periodically. The figures used in any budget must be confirmed directly with the chosen entity (for GS1, with the national member organisation in the country of the legal entity) before the budget is signed off. Numbers copied from third-party articles, including this one, are not a substitute for the entity's current published schedule.

This post is part of the EUDAMED, UDI and Registration category in the Subtract to Ship: MDR blog. Authored by Felix Lenhard and Tibor Zechmeister. For the upstream entity decision that drives most of the cost shape, read this post alongside how to obtain a UDI. For the label-side cost drivers, read UDI carrier requirements.