European defence tech has gone from an afterthought to one of the fastest-growing VC categories on the continent. That's good news. The less good news: the playbook you'd use for a SaaS seed round doesn't translate. This is a different game, and investors know it.
First: the market you're operating in
The macro context matters here because it shapes investor psychology. This sector was nearly invisible to VC five years ago. It isn't anymore.
Important context: This is patient capital. Investors entering defence tech know exits take longer — M&A by defence primes or public markets, not a quick trade sale. If your pitch implies SaaS-speed liquidity, that's a red flag on your side of the table.
The round: what seed looks like
Seed rounds in European defence tech are now well-defined. Generalist VCs have entered alongside specialists, and the NATO Innovation Fund and sovereign vehicles increasingly anchor early rounds.
A crucial structural difference: most European defence startups combine equity with non-dilutive capital — EU grants (EDF, Horizon Europe, Defence Equity Facility), national innovation agency co-funding, and pilot contracts that effectively subsidise R&D. If you're not pursuing this, you're leaving money on the table and signalling naivety to investors.
Team: what investors want to see
Team composition is screened harder here than in most sectors. A strong technical co-founder is necessary but nowhere near sufficient.
- Domain expert with defence/security background — ex-military, intelligence, or defence contractor who understands procurement constraints, doctrine, and operational realities
- Strong technical founder — software, AI, autonomy, sensing, or hardware with a track record of shipping robust, secure systems (not just demos)
- At least one person who can navigate government sales — tendering, RFIs/RFPs, security clearance requirements; this is a full-time expertise, not a side task for a founder
- Advisory board with ex-officers or defence-policy experts — not decorative; these relationships open doors that cold outreach can't
No domain expertise in the founding team? No access to security-cleared advisers? This is the fastest path to a "no." Investors won't teach you how the customer thinks — that's your job before you walk in.
Headcount
Typical seed-stage defence startup runs 5–20 FTEs: multiple engineers in autonomy, AI, embedded systems or cyber, plus at least one BD person with government relationships. Lean and focused, not sprawling.
Product & technical maturity
Investors think in TRL (Technology Readiness Level). You should too. The seed bar isn't "working prototype" — it's prototypes validated in relevant environments.
TRL expectations at seed
Software-heavy plays (C2, targeting, logistics, cyber)
- Production-grade software that integrates with existing military systems — specific data links, security baselines, relevant standards
- Demonstrated interoperability with at least one major platform or network
- Ability to operate at required classification levels
Hardware / autonomy (drones, robotics, sensing)
- Prototype systems operating in representative environments with clear, documented performance metrics
- Test reports or exercise feedback from actual users, not just internal evaluation
- Credible roadmap from current TRL to fielded system, with milestone clarity
Traction: it doesn't look like SaaS ARR
This is the biggest mindset shift for founders coming from software. "Traction" in defence tech is a different set of signals — and VCs are specifically trained to read them.
Defence ARR lags technical progress. Modest ARR at seed is normal and expected. What investors are actually reading is your pipeline, your contract trajectory, and your procurement access. A €200K ARR number means nothing without context on what's behind it.
What seed traction looks like
What else investors want to see
- Paid pilot contracts or tech evaluation programmes with defence or security agencies — not just MoUs or letters of intent
- Visible pipeline of additional pilots — shortlisted tenders, active RFI/RFP processes, or MoUs with named organisations
- Dual-use commercial revenue in adjacent markets (infrastructure inspection, border surveillance, maritime logistics, enterprise cyber) — this provides near-term cash flow and validates the technology outside the procurement bottleneck
- Acceptance into NATO DIANA, national defence accelerators, or EU defence innovation schemes — treated as quality signals, not just résumé items
Defensibility: what your moat looks like
The defensibility question in defence tech is different from software. It's not primarily about data flywheels or API integrations. It's about:
- Proprietary algorithms or architectures for targeting, perception, EW resistance, or secure communications — something that can't be replicated by bolting together open-source components
- Access to military-grade or sensitive datasets, with a clear legal and ethical framework — the data access matters as much as the IP
- IP strategy that survives export control and classification scrutiny — ITAR, dual-use regulations, end-user vetting; if you haven't mapped this, do it now
- Alignment with NATO/EU capability priorities — awareness and decision-making, freedom of manoeuvre, critical infrastructure resilience; if you can't locate your product on this map, neither can investors
Red flags that will kill your round
- No domain expertise in the founding team and no security-cleared advisers in the network
- No evidence of real buyer interest — no pilots, no sandbox trials, no active tenders being tracked
- No clear dual-use or commercial revenue path alongside the defence play — pure defence-only strategies are hard to scale at venture pace
- Overly optimistic timelines to multi-year contracts, with no recognition of how procurement actually works
- No clarity on which NATO/EU capability gap you're addressing — "defence AI" is not a positioning
- No understanding of ITAR, export controls, or dual-use regulations — this signals existential legal risk
- Heavy dependence on classified data with no plan for development and testing on unclassified or synthetic datasets
Quick benchmark: seed readiness
| Signal | Competitive at seed | Red zone |
|---|---|---|
| TRL | 4–5 entering; roadmap to 6 | Below TRL 3 with no clear path |
| Paid pilots | 1–2 with defence/security agencies | Zero; letters of intent only |
| ARR | Low–mid €100Ks (not the lead metric) | Zero with no pipeline story |
| Pipeline | €100Ks–few €Ms contracted or highly probable | No active tenders or RFPs tracked |
| Team | Domain expert + tech founder + gov BD | Engineers only, no defence access |
| Non-dilutive capital | Grants + pilot contracts layered in | Equity-only, no EU/national instruments pursued |
| Regulatory literacy | ITAR/export controls mapped; IP strategy clear | "We'll figure it out later" |
| Typical round | €2–5M equity · €8–15M pre-money · 24–30 months runway target | |
The honest framing: Defence tech VCs are betting on 10+ year timelines. What they need from you at seed isn't growth metrics — it's conviction that you can navigate a procurement system designed to resist you, build something genuinely hard to replicate, and find dual-use revenue to bridge the gap. If your story doesn't address all three, the round will be harder than it needs to be.
