Business & Strategy – Digital Health with Market View

By Christoph Eberhardt · CEO, DUX Healthcare · 19 April 2026

This hub is the business- and strategy-perspective on digital therapeutics in Germany and the surrounding European market — written for international investors, business-development leaders, founders evaluating Germany as a first reimbursement market, and DTx companies expanding into Germany from another jurisdiction. Market figures come from the GKV-Spitzenverband DiGA Reports and the SVDGV DiGA Reports and are consistently attributed to one of the two; where they diverge in interpretation, we mark the framing. DUX-own observations from 7 years of Software-as-Medical-Device development with five DiGAs (Digitale Gesundheitsanwendungen — Germany’s certified digital health applications) under contract on the mHealth Suite are flagged as DUX position.

What “Business & Strategy” covers in digital health

Digital health has two distinct literatures. The regulatory literature (BfArM, BSI, MDCG, FDA guidance) explains how to build a compliant product. The business literature – when it exists – explains whether to build it at all, on what terms, and how to finance the gap between the engineering spend and the first cash flow. This knowledge area is the second literature, applied to the European regulated-path reality from a manufacturer perspective.

Five question clusters structure this knowledge area:

1. Market truth. What is the DiGA market actually paying, growing, and consolidating at – read through both the GKV-Spitzenverband (the National Association of Statutory Health Insurance Funds – payer side) and the SVDGV (Spitzenverband Digitale Gesundheitsversorgung – the digital-health industry association) with explicit dual attribution.
2. Cost and pricing mechanics. List price, negotiated price, the 2.4× delta, the 20 % outcomes-linked rule from 2026, and what this does to the unit economics of a single-product DiGA company.
3. Operating-model choice. Build-vs-platform is not a philosophical debate in digital health – it is a 6–12× margin-compression decision over the mandatory-update calendar.
4. Go-to-market. Realistic timelines from project start to first prescribed unit, the legal-manufacturer gate for non-EU entrants, indication-selection economics, and the commercial logic of Class IIa vs. Class IIb.
5. Adjacent models. Pharma companion apps, orphan-drug companion economics, partnership structures with statutory health insurers under § 140a SGB V, and the exit landscape for listed-DiGA assets.

A deliberate exclusion: this knowledge area does not cover wellness apps, non-regulated DTx, consumer-health brands. The economics of the unregulated path are interesting, but they belong in a different playbook and a different cost of capital.

The 2026 market – what the numbers really show

Five full years of DiGA data produce a statistically meaningful dataset. The full analysis lives in the DiGA Market Report 2026 – GKV and SVDGV Dual-Check. The headline:

Both sources agree at the 31 December 2025 cutoff: ~1.6 million activation codes redeemed since September 2020; 58 DiGAs currently listed (48 permanent, 10 under Erprobung – trial); 74 ever-listed; 16 delisted; ~EUR 400M cumulative statutory spend of which EUR 170M fell in 2025 alone (GKV-Spitzenverband DiGA-Bericht 2025; SVDGV DiGA-Report 2025↗). The year-on-year 2025 growth rate is +63 % (GKV) / +64 % (SVDGV).

The two sources diverge on interpretation, and this is the single most important thing an international reader needs to internalise about the DiGA market: there is no neutral aggregator. The GKV-Spitzenverband files its report under a statutory mandate (§ 33a para. 6 SGB V) and argues for cost containment. The SVDGV is a private-law industry association representing 170+ e-health companies and argues for market expansion. Both publish verifiable figures. Neither is a neutral reference.

Three structurally important numbers for an investor thesis:

• The pricing delta. Per GKV-Spitzenverband, the average manufacturer list price is EUR 544; the average negotiated price (Vergütungsbetrag) after month 13 is EUR 227 – a 59 % reduction. The delta is larger in individual cases: levidex, a Class IIb digital therapeutic for multiple sclerosis, went from EUR 2,077.40 list to EUR 247.81 negotiated (-88 %) after a 29-month trial. For any DCF running on a DiGA asset, this single month-13 repricing event is the biggest commercial-model discontinuity in the lifecycle.
• Concentration risk. Per GKV-Spitzenverband Executive Summary, metabolic diseases account for 37 % of 2025 redemptions from 8 of 58 listed DiGAs. Psychological indications are 55 % of the directory but 27 % of redemptions. A handful of products carry a disproportionate share of the volume curve.
• Attrition signal. 9 manufacturer insolvencies since 2020, 16 DiGAs delisted (22 % of everything ever listed), 6 new listings in 2025 vs. 7 delistings (per GKV-Spitzenverband). Listing velocity is slowing; delisting velocity is accelerating. The supply side is consolidating even as demand grows.

The SVDGV frame on the same data is not wrong – it is differently oriented: 34 of 50 provisionally listed DiGAs successfully converted to permanent listing (~70 % conversion rate), which is a functioning fast-track by any reasonable benchmark. Both framings inform the investment decision; neither alone is sufficient.

Practitioner reading: the DiGA market is entering its first genuine consolidation phase. Growth is real and accelerating; entry barriers are real and rising; the 2026 outcomes-linked rule will widen the gap between platform-engineered and project-built operators.

Build vs. platform – the structural decision

This is the business-model decision that most founders underestimate and most investors ignore. The public regulatory literature is written as if every DiGA is bespoke. The internal economics say otherwise.

A custom DiGA build, measured honestly, costs €500k–€2M to bring to BfArM listing. The evidence study is the largest single line item at €200k–€1M+. The regulatory tracks (MDR Class IIa conformity assessment with Notified Body, ISO 13485 QMS, IEC 62304 lifecycle documentation, clinical evaluation, usability engineering under IEC 62366) absorb 3–12 months of engineering capacity. The data-protection and security tracks (ISO 27001 ISMS certification, BfArM’s ~150 data-protection criteria, BSI TR-03161 cybersecurity certification per DiGA, GDPR DPIA, penetration testing) add 2–4 months only if run in parallel with product work. Sequentially, the same tracks eat a year.

That is the first build. Every listed DiGA then absorbs a mandatory-update cadence that makes the private P&L decisively non-linear:

• 1 April 2022: ISMS certification to ISO 27001 or ISO 27001-on-IT-Grundschutz, accredited.
• 1 January 2024: interoperable export interface and ePA (elektronische Patientenakte – Germany’s electronic health record) connection under § 6a DiGAV; GesundheitsID under § 291 para. 8 SGB V.
• 1 January 2025: BSI TR-03161 per DiGA, application-relevant from 1 July 2025.
• 1 January 2026: ≥ 20 % success-dependent price share.
• 1 July 2026: AbEM stage I data collection (usage, drop-out).
• 15 April 2027: first AbEM submission to BfArM.
• 1 July 2027: AbEM stage II (PGI-C, patient satisfaction).
• 1 July 2028: AbEM stage III (indication-specific PROMs).

A custom-build operator absorbs each of these as a project release. A platform operator absorbs them once, across the portfolio, as shared infrastructure. Across seven mandatory-update events in six years, that is a structural 6–12× capacity multiplier on the compliance line – which translates directly into the ability to redeploy engineering capacity on AbEM-relevant outcome improvements instead of on rebuilding the change-management process for the fourth time.

DUX reading: for a single-product DiGA company with prior medical-device software experience, the build decision may be defensible. For a multi-product roadmap, a pharma companion portfolio, or any operator intending to list more than one DiGA across five years, the platform decision is commercially obvious – and sequentially inverted: the platform work happens before the DiGA project, not inside it.

Go-to-market timing – realistic cycles

The single most frequent international misreading of Germany is the “three months at BfArM” headline. The statutory clock in DiGAV § 16 para. 1↗ is three months from the point of administrative completeness – and administrative completeness requires 26 content blocks at filing (DiGAV § 2 para. 1), most of which are neither administrative nor trivial: the clinical evaluation, the CE certificate under MDR, the BSI TR-03161 data-security certificate, the penetration test report, the GDPR impact assessment, the study or evaluation concept, the ePA interface specification, the AbEM integration, and the price declaration.

A realistic calendar from project initiation to first prescribed unit, for an international team starting from zero on German regulatory specifics:

• Ramp-up (intended use, evidence strategy, regulatory path, legal-manufacturer arrangement): 3–9 months. For non-EU teams, dominated by the legal-manufacturer decision.
• Product and regulatory work (MDR conformity assessment, IEC 62304, ISO 13485, risk management, clinical evaluation, usability): 3–12 months, class-dependent, Notified-Body-dependent for Class IIa and above.
• Data protection and security (BSI TR-03161 per DiGA, ISMS, BfArM criteria, GDPR DPIA): 2–4 months if run in parallel. Sequential: a year.
• Pivotal comparative study: 6–18 months for Class I/IIa during trial listing; must be complete pre-filing for Class IIb.
• BfArM filing, formal review, decision: 3–6 months, completeness-dependent.

For a platform-engineered operator, Phase 1 compresses to twelve weeks until a DiGAV-compliant, CE-marked product, provided intended use is defined and the regulatory posture maps onto the platform’s existing certifications. The legal-manufacturer and study slices are not platform-compressible – they are timeline-compressible only by solving them in parallel.

The operational implication: product launch timing is dominated by the evidence study and the legal-manufacturer setup, not by the BfArM clock. Investor memos that frame Germany as a “three-month reimbursement win” are either reading the statute wrong or describing a project that is three months away from listing after twelve to eighteen months of invisible pre-work.

The Class IIb wrinkle. Since Germany’s 2024 DigiG law, Class IIb DiGAs are eligible – but the preliminary-listing route is closed to them (DiGA Guide v3.6 Sec. 2.3.1). A Class IIb product must demonstrate medizinischer Nutzen (medical benefit) at filing, not just a structural or procedural improvement. In commercial terms, this means the pivotal study sits before reimbursement starts, which re-orders the financing curve dramatically – no bridge-financed list-price year one, no revenue until listing, and a study that typically runs 12–24 months on its own.

Success and failure patterns

This section is opinionated. The patterns below come from DUX mandate experience and the publicly documented market record (insolvencies, delistings, GKV-Spitzenverband annual reports) – not from a representative survey. Where DUX takes a position, it is marked.

Failure patterns that recur

1. Intended use written by product, not by regulatory. A DiGA’s intended use is the contract with BfArM, with the evidence study, and with the pricing negotiation. Intended-use statements drafted to “match the roadmap” rather than to match the indication and comparator produce studies that do not transfer to real-world patient populations, pricing bands that do not support the cash-burn curve, and listings that restrict the population in ways marketing has not anticipated. Roughly one in five trial-to-permanent transitions lists with restrictions, per GKV-Spitzenverband 2025.

2. Security retrofitted, not engineered. Teams that treat BSI TR-03161 as a final check before filing rebuild half the backend architecture. The cost balloon is not in the certification fee but in the rebuild. Platform operators carry shared identity, segmentation, and logging layers forward; project operators pay the rebuild cost per product.

3. Evidence study run in theory, not in reality. “Usual care” in an academic trial is not the status quo of average German primary care. Underpowered samples, comparators chosen for convenience, and protocols written without pre-specification produce trial results that are formally legal but commercially unusable. A failed pivotal triggers de-listing, a 12-month lockout, and no second preliminary listing (DiGAV § 16).

4. Cash-burn clocks misaligned with the 13-month negotiation. A manufacturer that priced the list-price year at €1,500 and modelled the business case on that figure is modelling a 12-month revenue line, not a steady state. The 59 % price drop at month 13 is the single biggest discontinuity in the lifecycle. Nine insolvencies since 2020 are a reminder that this discontinuity is not theoretical.

5. Legal-manufacturer decision solved last. For non-EU companies, waiting until month ten to set up the MDR legal manufacturer typically means delaying the BfArM filing by nine months. The structural fix is solving the legal-manufacturer question in month one – subsidiary, partnership, or acquisition – and building the rest of the roadmap around that decision, not the reverse.

Success patterns that recur

• Indication chosen for evidence, not for TAM. The DiGAs that convert from trial to permanent listing and survive the pricing negotiation tend to be in indications with well-validated PROMs, clear comparators, and clinically meaningful minimum differences that can be powered in reasonable sample sizes. Psychotherapy-adjacent indications have been disproportionately represented in successful conversions – partly because CBT protocols offer strong comparator logic.
• Platform economics assumed from day one. The operators publicly profitable in year four have either platform-built or platform-bought. The ones still building custom in year four typically need another funding round to cover the next mandatory-update cycle.
• Pharma partnership as a second leg. Digital therapeutics operators who add a pharma-companion leg (not as a pivot, but as a parallel revenue line) convert engineering overhead into a second distribution channel without repeating the evidence stack.
• Legal-manufacturer solved as a gating step. Teams that accept month-one that the legal manufacturer is a prerequisite, not a workaround, tend to have the rest of the timeline behave rationally.

DUX reading: most DiGA failures are process failures, not evidence failures. Evidence failure is detected late (after study completion) and is survivable only by running a second study, which most operators cannot finance. Process failure is detectable early enough to correct – if the team has the regulatory depth to read the signs.

Pharma companion and strategic partnership models

Outside the pure DiGA track, two adjacent commercial models deserve attention in any European digital-health thesis.

Pharma companion apps. A companion app in the European regulatory frame is a medical device under MDR, typically Class IIa, that supports the use of a prescription drug – titration, adherence, symptom-tracking linked to dosing decisions, outcome reporting. The commercial logic is not reimbursement; it is either co-funded with the pharma sponsor as part of a therapeutic package, or listed as a DiGA where the indication and evidence frame allows. The difference from a pure DiGA is structural:

• Revenue model: pharma co-funding or milestone-based payment, not statutory reimbursement. Product P&L is a cost centre; sponsor ROI is tied to drug sales and adherence curves.
• Evidence bar: often higher than DiGA baseline, because the pharma sponsor’s regulatory file may treat the app as a co-determinant of drug efficacy.
• Time-to-market: comparable or slightly faster than DiGA if the pharma partner has an existing CE-marked platform, significantly slower if the app must be built from scratch.
• Commercial exit: pharma partnerships do not have a “listed asset” exit. They have a contract renewal conversation every 2–3 years.

The model is structurally interesting for orphan drug companion apps – an underserved segment where small patient populations make standalone DiGA economics weak but make pharma-sponsored companion economics viable.

Selective contracts under § 140a SGB V. A parallel path for digital health products in Germany is direct contracting with individual statutory health insurers (Krankenkassen) under § 140a – the “special care” clause. A selective contract does not require BfArM listing, is negotiated bilaterally with one or a consortium of insurers, and covers a defined regional or population segment. The trade-off: smaller reach (one fund covers 1–10 million lives vs. 73 million via DiGA), higher commercial flexibility, no AbEM exposure. Selective contracts are underrated in the English-language literature – they work for specialist indications where the DiGA evidence bar is a 24-month obstacle but a willing fund partner is a 6-month decision.

Innovation Fund (Innovationsfonds). Germany’s § 92a SGB V Innovation Fund finances evaluation projects for new care models, often including digital components. It is not a revenue channel – it is non-dilutive pilot-financing that can bridge 18–36 months for operators without listed-product revenue, provided they read it as evaluation-first, not commercialisation-first.

Where DUX fits – platform economics in practice

DUX Healthcare does not sell custom DiGA development. The model is platform-engineered DiGAs on the mHealth Suite – 80+ pre-validated modules where MDR documentation, BSI TR-03161 architecture, data-protection design, the ePA interface, and AbEM data capture are shared infrastructure rather than per-project rebuilds. Five DiGAs currently run on this platform, each individually tested against BSI TR-03161 per DiGA, all CE-marked under MDR. At the process level, the software development lifecycle is certified under BSI TR-03185 – Secure Software Lifecycle; DUX is the first company to hold that certification. Platform work happens before a DiGA project, not inside it, which is what makes the twelve-week DiGAV-compliant, CE-marked product real rather than aspirational.

For international teams the platform carries a second structural advantage: the MDR legal-manufacturer role through a partner network, so the two heaviest market-entry blockers (DiGAV engineering depth and the EU-manufacturer requirement) are solvable in parallel rather than serially.

The commercial mechanics of the platform sit at /build/ for operators considering a DiGA build. The investor track sits at /investors/. Both bridges are deliberately separate from this hub: this hub exists to inform, not to sell.

Frequently asked questions