In the startup landscape, innovation is the rule, not the exception. Some innovations concern software, data, hardware, or business models. Others, however, are radically different: they are based on scientific discoveries and frontier technologies. These are deeptech startups, and they follow different rules compared to all others.

Market-driven vs Technology-driven

Deeptech startups are technology-driven. While this may seem obvious, it fundamentally changes their path compared to traditional startups.

In traditional startups, everything starts from a concrete problem: you build an MVP, test it in the market, collect feedback, and iterate until you find product-market fit. Technology is only a tool to serve the customer.

In deeptech startups, the path is reversed. You start from a discovery or technological innovation — an algorithm, a material, a process, or a physical component — and first demonstrate that it works. Only afterward do you explore the markets where it might have value. The product development follows the technology, not market demand.

This has concrete consequences: the priority is not rapid market testing, but reducing scientific uncertainty. This often means years of research, expensive prototypes, and long timelines to bring the technology to a sufficiently mature stage.

This does not mean that deep tech startups lack a business model; having one is essential for raising capital.
Instead, their business logic is usually long-term oriented and evolves alongside technological development, rather than being validated immediately at the earliest phases.

Specifically, traditional startups usually follow these development stages:

1. MVP: Demonstrates that the identified problem truly exists, is felt by users, and that they are willing to spend time or money to solve it.
2. Early adopters: Shows that a small group of users loves the solution and uses it repeatedly and intensively.
3. Product-market fit (PMF): Demonstrates that the product has a real and scalable market.

In deeptech startups, the life cycle is dictated by the TRL (Technology Readiness Level), which indicates the stage of technology development and identifies remaining phases to achieve maturity and commercialization:

• TRL 1–3: Research & Proof of Concept – Basic scientific exploration, applied research, and experimental validation of feasibility in the laboratory environment.
• TRL 4–6: Prototyping & Validation – Laboratory testing, simulation in realistic but controlled conditions, and real-world prototype demonstration.
• TRL 7–8: Pilot & Operational Deployment – Initial small-scale use and large-scale operational implementation.
• TRL 9: Market Ready – Fully mature, commercializable technology.

Advancing through the TRLs is a long, step-by-step process, reflecting the years of research, prototyping, and validatioon required before the technology can reach the market.

Speed vs Patience

Traditional startups are usually focused on speed, moving from zero to MVP in a few weeks and to a finished product in a few months.

Deeptech startups are different. Developing and validating a science-based innovation to the point of commercialization requires more time, resources, and specialized talent. If the invention involves creating a new substance or process, additional time is needed to develop a manufacturing process capable of producing it at scale.

If this new manufacturing process is unlike anything existing manufacturers know, the startup may need to build its own production facility, which is capital-intensive and time-consuming.

For all these reasons and more, deeptech startups generally take longer to reach the market and require significantly more capital than traditional startups.

This also affects funding choices: many VCs have a 10-year horizon, expecting at least a 10X return in that time — often too short for taking a scientific innovation from zero to large-scale commercialization.

While traditional startups can almost always raise VC funding, deeptech startups are often initially funded by public grants, either through dedicated programs or university research (delaying prototype commercialization to maximize available resources).

Obtaining grants in deep tech is not about presenting a finished product, but about structuring technological uncertainty in a credible way.

First, the startup must precisely define what is still unknown from a scientific perspective, and why this uncertainty cannot be addressed through market mechanisms or private capital. Grant evaluators fund unresolved scientific or technological questions, not execution risk.

Second, the proposal must show that this uncertainty is reducible. This means formulating testable hypotheses, defining experimental approaches, setting measurable technical milestones, and showing how the funds will contribute to advancing the technology along the TRL scale. The goal is not to claim success in advance, but to demonstrate a rigorous path toward validation.

Third, the proposal should not obscure potential negative outcomes. Instead, it should anticipate their underlying causes, explain how these risks can be mitigated, and outline credible alternative paths in the event of failure. In deep tech, acknowledging failure scenarios is often a signal of technical maturity.

Public funding offers advantages such as:

• No repayment required
• No dilution of equity
• Quality assurance

But it also comes with disadvantages:

• Often only covers R&D costs, not marketing
• Selection criteria can be highly specific and tied to government interests, not entirely dependent on the startup
• May require predefined milestones, assuming an already advanced project stage
• Restrictions on how money is used, potentially impacting research
• Bureaucracy, which slows down the process

Fortunately, the number of VCs specialized in vertical deeptech sectors — such as space, clean energy, and AI — has been rapidly growing in recent years.

Pitch like an engineer

Whether applying for public grants or raising funds from specialized VCs, you will need to prepare a project presentation. Often, even public grants require an oral pitch, usually after an initial selection.

In a deeptech pitch, there are many differences compared to a traditional startup. Beyond the fundamental and obvious elements, there are other specific characteristics to highlight.

• Why now

The outcome of research does not depend solely on the team’s skills, but largely on the technological and scientific context. Ignoring this is one of the gravest mistakes a deeptech startup can make.

In deeptech, “why now” demonstrates that recent technological, scientific, or cost breakthroughs have finally made previously unsolvable or economically unviable problems achievable. Drastic reductions in compute costs, new materials, AI advances, or regulatory changes can turn historically prohibitive challenges — such as drug discovery or industrializing new technologies — into concrete, scalable opportunities.

This slide must show that the idea is not just new, but perfectly timed: it emerges at a unique moment when technology, market, and regulation align to open a window of opportunity that did not exist before and will not remain open for long.

A convincing “why now” eliminates the risk of being too early or too late and signals to investors that if they do not act now, they will miss a once-in-a-lifetime opportunity.

• Team

The team slide is essential for any pre-seed startup, as VCs at this stage primarily evaluate the people. This is even more critical in deeptech.

Scientific innovation involves complex, long-term challenges requiring deep, often heterogeneous technical expertise. It is not enough to be competent; founders must demonstrate excellence, or at least be on the trajectory toward it.

Excellence does not mean being a world-class engineer or coming from elite research centers (e.g., CERN). What matters is showing that the founders’ life trajectories naturally and rapidly converge toward that level. Evidence can include:

• Years of obsessive work on the same problem
• Personal connections with key experts, granting access to information difficult to obtain otherwise
• Prior prototypes, systems, or solutions similar to what is currently being developed

In a well-founded deeptech startup, the company’s outcome should appear inevitable: not the result of opportunistic ideas, but the natural culmination of a life dedicated to solving that specific problem.

• Go-to-market

This slide must clarify how the technology moves from lab to market. In deeptech, even excellent innovation rarely succeeds without a clear path to customer value.

It is essential to demonstrate understanding of the entire adoption process: sales timelines, customer education, regulatory requirements, and scaling strategy (channels, local partnerships, OEMs, or system integrators).

The revenue model must be clearly stated (hardware + software, licenses, recurring revenue, consumables, or services), indicating volume thresholds for profitability and distinguishing what scales from what remains structurally costly. Key supply chain risks and design-for-manufacturing choices should also be addressed.

Finally, show traction and validation, both commercial (pilots, early customers) and technical (IP, approvals, strategic partnerships), to demonstrate that the technology is adoptable, scalable, and defensible.

Pro tip: Deeptech founder is ready to pitch only when they can explain their product both to a 12-year-old and to a PhD.

Author: Marco Carabelli