How to Evaluate a Deep Tech Startup

This guide gives you a practical, step-by-step framework for evaluating a deep tech startup—whether you’re considering investing, joining as an employee, or partnering. It focuses on the unique risks of deep tech: long development timelines, scientific uncertainty, capital intensity, and the need for specialized talent. You’ll learn what to check, what questions to ask, and how to spot red flags early.

What You’ll Need

Access to the startup’s pitch deck, financial model, and any technical documentation (e.g., patents, whitepapers, lab results).
Basic understanding of the technology domain (or a trusted technical advisor).
Time to conduct interviews with the founders, team members, and ideally a few customers or partners.
A list of comparable companies or technologies for benchmarking.
Ability to do background checks on founders and key hires (LinkedIn, academic publications, prior startups).

Step-by-Step

Define your investment or employment thesis. Before looking at any startup, write down what you’re looking for: stage (seed, Series A), technology area, risk tolerance, expected return or career goals, and time horizon. This helps you filter opportunities quickly and avoid falling for a shiny but misaligned pitch.
Evaluate the problem and market size. Deep tech often creates new markets rather than serving existing ones. Ask: What is the specific problem being solved? Who has it today? Is there a regulatory or economic push that makes this problem urgent? Look for a bottom-up market estimate (number of potential customers × price) rather than a top-down “total addressable market” that assumes everyone will buy.
Scrutinize the technology and IP. Read the patents (if any) and check their claims, not just the title. Who owns the IP? Are there licensing agreements? Is the core technology protected or easily replicated? For unpatented tech, assess the “know-how” barrier: how hard is it for a competitor to reverse-engineer? Also check for freedom-to-operate issues—does the startup infringe on existing patents?
Assess the scientific and engineering risk. Deep tech often has a “valley of death” between lab demo and commercial product. Ask the team: What are the top three technical risks? What experiments have been done to de-risk them? What is the plan for scaling from prototype to production? Look for evidence of third-party validation (e.g., peer-reviewed papers, independent lab tests) rather than only internal claims.
Review the team’s capabilities and track record. Deep tech requires a rare combination of scientific depth and business execution. Check: Do the founders have domain expertise? Have they built companies before? Is there a strong CTO or chief scientist? What about the rest of the team—are they experienced or fresh graduates? Look for complementary skills: a visionary scientist paired with a pragmatic operator is ideal.
Analyze the business model and go-to-market. How will the startup make money? Is it a product sale, licensing, or service? Who is the customer—enterprise, government, consumer? Deep tech often requires long sales cycles; ask about pilot programs, letters of intent, or early revenue. Be wary of a business model that relies on a single large customer or a government grant that may not recur.
Examine the financials and burn rate. Even early-stage startups should have a clear budget. Ask for a monthly burn rate, current runway, and a breakdown of spending (R&D vs. sales vs. ops). Compare the burn to the stage: a seed-stage company burning $500k/month with no revenue is risky unless they have a clear path to Series A. Also check the cap table: who owns what? Are there any unusual liquidation preferences or anti-dilution clauses that could hurt you as an investor or employee?
Talk to customers (or potential customers). If the startup has any users, speak with them directly. Ask: What problem does this solve for you? Would you be upset if it disappeared? How much would you pay? For pre-revenue startups, talk to potential customers in the target market to gauge interest. Be skeptical of “we talked to 50 customers and they all loved it”—ask for specifics.
Check the competitive landscape. Who else is working on the same problem? Is the startup’s approach fundamentally different or just a minor improvement? Look for moats: proprietary data, network effects, regulatory approvals, or manufacturing know-how. If the only advantage is “we’re smarter,” that’s weak.
Evaluate the regulatory and ethical risks. Deep tech often intersects with regulation (e.g., medical devices, autonomous vehicles, AI). Is the startup aware of the regulatory path? Have they engaged with regulators? What about ethical concerns—could the technology be misused? These risks can kill a company regardless of technical success.
Do background checks on founders and key team. Search for any history of fraud, lawsuits, or failed startups. Also check their academic and professional history for exaggerations. A pattern of overclaiming is a huge red flag.
Negotiate your deal or role with clear metrics. If you decide to invest or join, define clear milestones. For investors: what are the next funding triggers? For employees: what is the equity vesting schedule, and what are the performance expectations? Make sure your compensation reflects the risk—deep tech is high-risk, so you should be rewarded accordingly.

Common Mistakes

Falling for the “science is cool” trap. Just because the technology is impressive doesn’t mean it’s a good business. Always separate the science from the commercial viability.
Ignoring the “valley of death.” Many deep tech startups die between prototype and production. Ask specifically about manufacturing scale-up and cost reduction plans.
Overvaluing patents. Patents are only as good as their claims and enforceability. A patent on a narrow improvement may not block competitors.
Underestimating the timeline. Deep tech often takes 5-10 years to reach profitability. If you need a quick return, look elsewhere.
Trusting the financial model. Early-stage financial models are often overly optimistic. Stress-test the assumptions: what if revenue is half of what’s projected? What if the burn doubles?
Not talking to the technical team. Founders may be great salespeople, but the actual engineers and scientists can give you a clearer picture of the technology’s readiness.

Tips

Use a scorecard. Create a weighted scorecard for criteria like technology, team, market, and business model. Score each startup consistently to compare opportunities.
Look for a “technical advisory board.” A strong advisory board of renowned scientists or engineers adds credibility and can help de-risk the technology.
Ask about “killer experiments.” What single experiment would prove or disprove the core hypothesis? If the team hasn’t done it yet, that’s a risk.
Check for government grants or non-dilutive funding. These can extend runway and provide external validation (e.g., SBIR, Horizon 2020).
Consider the exit path. For investors: who might acquire this company? For employees: is there a realistic IPO or acquisition path that makes your equity worth something?

Troubleshooting

What if the technology is too complex for you to understand? Bring in a technical advisor or do a deep dive with the CTO. Ask them to explain the technology in simple terms and then ask probing questions. If they can’t explain it clearly, that’s a red flag.

What if the team is evasive about financials? This is a major warning sign. Insist on seeing the cap table and burn rate. If they refuse, walk away—transparency is essential in deep tech.

What if the market seems too small? Ask about adjacent markets or pivots. Sometimes a deep tech platform can be applied to multiple industries. But be cautious: a pivot is risky and may indicate the original thesis was flawed.

What if you find a competitor with a similar technology? Compare the teams, IP, and execution speed. The first to market isn’t always the winner, but a fast follower with better resources can overtake.

What if the startup is pre-revenue and has no customers? Focus on the quality of the advisory board, the strength of the IP, and the team’s track record. Pre-revenue is normal for deep tech, but there should be a clear plan to get to pilots.

Frequently Asked Questions

Q: How is deep tech different from regular tech startups?
A: Deep tech is based on scientific breakthroughs or engineering innovations that require significant R&D, long development cycles, and often regulatory approval. Regular tech startups typically build on existing technologies and scale faster.

Q: What is the most important factor to evaluate?
A: The team is often cited as the most important. A brilliant team can pivot, but a weak team can’t execute even a great idea. However, the technology itself must have a clear path to commercialization.

Q: How much time should I spend evaluating a deep tech startup?
A: Plan for 2-4 weeks of due diligence, including multiple interviews, technical review, and market analysis. Rushing can lead to costly mistakes.

Q: Can I evaluate a deep tech startup without a technical background?
A: Yes, but you’ll need to rely on trusted advisors or ask the startup to provide independent technical validation. Focus on the business model, team, and market, and bring in experts for the tech.

Q: What are common red flags?
A: Unrealistic timelines, over-reliance on a single person, lack of IP protection, vague answers about technical risks, and a business model that doesn’t clearly show how the technology becomes a product.

Q: Should I invest in deep tech as a beginner?
A: Deep tech is high-risk and requires patience. If you’re new to investing, consider diversifying across multiple startups or investing through a fund that specializes in deep tech.

General steps for current mainstream versions; menus can differ by device and version. See the disclaimer or report an error.

More in this bed

The whole Tech How-To bed →