Germany's IT skills gap is smaller than it was three years ago, and many founders read that as a signal that hiring will get easier. For a Series A or B company with a roadmap tied to a funding milestone, that conclusion is premature. The aggregate number has fallen, but the specialists a deep tech scaleup needs are still scarce, and the cost of a slipped quarter has not changed. The practical answer is to plan engineering capacity as a portfolio: permanent hires for the core, and a contracted, outcome-based layer for work that cannot wait for a recruiting cycle.
What the 2026 numbers actually say
Bitkom's labour market study for 2026, based on a survey of German companies conducted between calendar weeks 18 and 25, reports that 79,000 IT specialists are missing in Germany. The comparable figures were 109,000 unfilled IT positions in 2025 and 149,000 in 2023. On paper, the gap has roughly halved in three years.
The same study shows why that is not the end of the story:
- 78 percent of companies say they currently observe a shortage of IT specialists.
- 76 percent expect the shortage to worsen in the coming years. Only 8 percent expect it to ease.
- 49 percent of companies with IT positions expect AI adoption to increase demand for specialists with AI knowledge.
- 33 percent cite missing knowledge of current technologies as a hiring difficulty.
- 18 percent say they rely more on external IT specialists, and 6 percent recruit more from abroad.
Read together, these figures describe a market where general demand has cooled with the economy, while demand for current, specialised skills remains tight and is expected to tighten again.
The wider picture: demographics, capital and AI
The Bitkom figures sit inside a longer trend that will not reverse with the business cycle.
- Demographics. IW Köln estimated in 2024 that 16.5 million baby boomers will pass retirement age by 2036, while only about 12.5 million people of working age enter the labour market.
- Shortage lists. The Bundesagentur für Arbeit's bottleneck analysis for 2025 counts 157 shortage occupations, down from 163, and no longer reports a shortage for software development as a whole. That supports the point above: the general market has eased, the specialist market has not.
- Capital. According to EY's Startup-Barometer, German startups raised about EUR 8.4 billion in venture capital in 2025, 19 percent more than in 2024, across 716 rounds, the lowest count since 2019. Eighteen rounds exceeded EUR 100 million. Fewer, larger rounds mean higher milestone expectations per round.
- AI. Bitkom's 2026 AI survey finds that 57 percent of German companies with at least 20 employees use AI, up from 36 percent in 2025 and 20 percent in 2024. Demand is shifting toward engineers who can build and govern AI systems, which is precisely the scarce profile.
Why the aggregate gap is the wrong metric for a scaleup
A Mittelstand IT department hiring a system administrator and a robotics scaleup hiring an engineer who understands ROS 2, real-time constraints and safety cases are not in the same market. The Bitkom figure is an average across all company sizes and roles. Three effects make the specialist market tighter than the headline suggests:
- Skill specificity. AI/ML, embedded, medtech software under IEC 62304 and defence-adjacent work each draw on a small pool. A cooling general market does not enlarge that pool.
- Competition from larger employers. When large companies slow overall hiring, they usually keep hiring their scarcest profiles. Scaleups compete with them for exactly those people.
- Time, not just availability. A candidate who exists is not a candidate who starts. Notice periods of three months are common in Germany, and senior engineers often negotiate longer. Add sourcing, interviews and onboarding, and the first productive month can sit two quarters after the decision to hire.
Translating the gap into roadmap risk
Boards rarely ask how many engineers you hired. They ask whether the milestones the round was priced on will land. A useful exercise is to express hiring risk in milestone terms.
| Step | Question | Output |
|---|---|---|
| 1 | Which milestones in the next 18 months depend on roles you have not yet filled? | List of milestone and role pairs |
| 2 | For each role, what is your realistic time from approval to first productive month, based on your own last five hires? | Lead time in months |
| 3 | What happens to the milestone if that lead time doubles? | Slip in weeks per milestone |
| 4 | What does one month of slip cost in runway, revenue or valuation? | Cost of delay per month |
| 5 | Which parts of that work could be specified as deliverables with acceptance criteria? | Candidate scope for contracted delivery |
Illustrative calculation: assume a company burns EUR 400,000 per month and a product milestone that gates the next round slips by three months because two backend roles take longer to fill than planned. The direct cost is EUR 1.2 million of runway consumed without the milestone, before any effect on valuation. The numbers are hypothetical, but the structure holds for most funded companies: the cost of delay usually dwarfs the cost difference between hiring models.
A portfolio model for engineering capacity
The companies that handle the shortage well tend to separate three layers of work:
- Core layer. Architecture ownership, product direction, the domain knowledge that differentiates the company. This should be permanent, ideally local, and hired early even when slow.
- Delivery layer. Well-bounded workstreams with clear interfaces: a data pipeline, a test automation framework, an integration, a migration. This work can be specified, accepted and contracted.
- Elastic layer. Short peaks, specialist reviews, audits. Freelancers and boutique consultancies fit here, with attention to Scheinselbständigkeit risk under SGB IV §7 if the engagement starts to look like employment.
The delivery layer is where most roadmap risk sits and where most companies make structural mistakes. If external engineers are simply placed into the internal team and directed day to day, the arrangement can drift toward Arbeitnehmerüberlassung under AÜG §1, with the consequences described in §9 and §10 AÜG. If the work is specified as a Werkvertrag under BGB §631 with a Leistungsbeschreibung and formal Abnahme under BGB §640, the vendor carries delivery responsibility and directs its own people.
Questions to settle before the next hiring plan
- Which roles are truly core, and which are capacity you need for 12 to 24 months?
- Can you write acceptance criteria for the delivery-layer work today? If not, that is the first task.
- Who internally owns the interface to an external delivery team: a named product owner with time allocated, not a shared responsibility?
- What is your data and IP position if part of the work runs outside the EU? DSGVO Art. 44 ff. and Standardvertragsklauseln apply to transfers to countries such as India.
- What would you tell the board if a key hire falls through in month two?
The takeaway
A smaller national shortage is good news for the economy, not necessarily for your next milestone. Model hiring as a lead-time risk, put a price on delay, and keep the specified, contractible part of the roadmap independent from the recruiting funnel. The most important decision is to recognise that the delivery layer exists. Statements on AÜG and contract types in this article are general information, not legal advice.
The delivery layer only reduces roadmap risk if it is set up as delivery, not as headcount. niogod™ staffs that layer with a three-person pod under a Werkvertrag or Dienstvertrag with Abnahme per sprint, so the work is specified, steered by the vendor and accepted by you (see the pod structure). If your next funding milestone depends on roles you have not yet filled, that is a conversation worth having early.
Talk to niogod™
Planning engineering capacity in DACH? Book a 30-minute consultation with the niogod™ team.
Book consultation