In deep tech, the team isn't just important, it's often the only real asset at the seed stage. Robotics in particular demands a rare combination: deep engineering expertise, operational resilience, and the commercial instinct to actually sell something physical into conservative industrial environments.

We look at whether the founding team has walked this road before. Have they shipped hardware? Have they sold into enterprise? Do they know the difference between a research prototype and a production-ready system, and do they have the humility to close that gap?

A complete, complementary team with relevant domain experience and prior startup exposure scores high here. Serial founders with robotics-specific exits score highest. First-time founders without domain knowledge, however smart, start from a harder position. This criterion carries triple weight because, at seed stage, we're ultimately betting on people.

Where is the technology, really? We use the Technology Readiness Level (TRL) scale as a common language to assess how mature the core technology is. TRL 1–2 means it exists on paper or in simulation. TRL 7–8 means it's been demonstrated in a real-world environment, under real conditions, with real edge cases.

For a seed investment, we don't expect TRL 9, but we want to understand exactly where the startup sits and what it will take to get to the next level. Inflated TRL claims are a red flag. Honest, grounded self-assessment is a green one.

Robotics startups get copied. Hardware gets reverse-engineered. Software gets rebuilt. The question isn't whether a startup has a good product today, it's whether they'll still have a defensible position in three to five years, when the market heats up and better-funded competitors arrive.

We look at the full picture of defensibility: patents (granted, not just pending), proprietary software stacks, hardware-software integration that's genuinely hard to replicate, network effects, and data advantages that compound over time. Trade secrets alone rarely hold. A combination of IP, deep integration, and switching costs is where durable moats are built.

There's a world of difference between "customers are interested" and "customers are paying." We want to see evidence, not opinions. This criterion tracks the progression from first conversations to recurring revenue.

Early-stage signals like Letters of Intent or informal feedback are a start, but they're cheap to obtain. Paid pilots, signed contracts, and best of all, customers who keep buying and expanding their usage are what we're looking for. A clearly defined Ideal Customer Profile (ICP) matters too: founders who know exactly who they're selling to, and why, close deals faster and waste less runway.

Can this business actually scale, or will it always be a services company wearing a product company's clothes? Project-based revenue with long delivery timelines and thin margins is a warning sign in robotics.

We want to see a path to a repeatable, scalable model: standardized deployment, healthy unit economics, and sales cycles that shorten as the product matures. The best robotics businesses find ways to turn complex physical deployments into something closer to a software sale: predictable, recurring, and increasingly automated.

How long does it take for a customer to go from "yes" to actually using the product? In robotics, this question cuts to the heart of scalability.

A solution that requires a full facility redesign, months of custom engineering, and safety certification before a single robot moves is a fundamentally different business than one that ships in a box and is productive on day one. Shorter implementation times mean faster revenue recognition, lower customer acquisition costs, and a sales motion that can actually scale. Long, complex implementations compress margins, strain customer relationships, and make repeatability nearly impossible.

Annual Contract Value tells us whether the economics of this business can ever make sense at scale. A low ACV in a capital-intensive hardware business is almost always a structural problem, not one that sales velocity alone can fix.

We look at what customers are actually willing to pay today, and what the ceiling looks like as the product matures. High ACV markets give startups the runway to build proper sales teams, invest in customer success, and absorb the inevitable complexity of physical deployment. Low ACV markets demand a level of efficiency and standardization that most early-stage robotics companies simply can't achieve.

Can the startup sell the same product to the next customer, or does every deal require starting from scratch? Highly fragmented markets, where every customer has unique requirements, are brutal for robotics startups. Every customization eats engineering time, delays deployment, and destroys unit economics.

Markets with strong homogeneity, where customers share similar workflows, environments, and requirements, allow startups to build once and deploy many times. That's where product companies are made. This criterion rewards startups that have found (or are building toward) a segment where repeatability is genuinely possible.

Long sales cycles are a tax on startup survival. In robotics, where enterprise procurement processes, safety sign-offs, and pilot requirements are the norm, this tax can be existential.

We assess the typical time from first contact to signed contract in the target market. Multi-year tender processes are punishing for early-stage companies. Champion-led deals that close in months, or transactional models that close in weeks, are where startups can build momentum, iterate on feedback, and reach meaningful scale before running out of runway.

Is the market actually ready to buy, or is the startup still educating customers about why they need the product at all? Market readiness captures the demand-side dynamics: regulatory environment, cultural openness to automation, and the maturity of the customer's internal processes for evaluating and adopting new technology.

A market full of active inbound demand, where customers are competing to be early adopters, is a very different proposition from one where every conversation starts with overcoming skepticism. Startups in ready markets spend their energy on closing; startups in resistant markets spend it on convincing.

Ultimately, the size of the opportunity has to justify the risk, and the capital. We don't just look at today's market. We model where it's headed over a 10-year horizon, using a straightforward approach: current addressable customers × product price × compounded growth rate. This gives us a realistic view of the long-term upside, adjusted for how fast the market is actually expanding.

A market below €100M is unlikely to produce the outcomes that justify a venture investment. Markets above €10B, especially those at an inflection point driven by labor scarcity, regulatory tailwinds, or falling hardware costs, are where category-defining robotics companies get built. This criterion carries triple weight because market size ultimately determines the ceiling on what's possible.