Deep Tech Investment Criteria: What We Screen For Before We Invest
Most funds tell founders what they care about in adjectives. Ambitious teams. Breakthrough science. Real impact. None of it is falsifiable, which means none of it is useful to a founder deciding whether to spend three weeks preparing a data room for us.
So here is the version with numbers in it. These are the deep tech investment criteria we apply, the thresholds we look for, and the point at which we think a hard technology is ready to scale rather than ready to pitch. If you are building in deep tech and you read this and conclude you are two years early for us, that is a useful outcome. It saves you the three weeks.
A note on the odds before anything else. We evaluate roughly 3,000 companies per fund and invest in fewer than 25. That ratio is not a filter we are proud of; it is a consequence of holding reserves for follow-on capital in the companies that work, which means fewer first cheques. Read the rest of this with that number in mind.
We fund technologies that create their own markets
The first thing we assess is what I call systemic breakthrough potential — technologies that fundamentally reimagine how an industrial process, an energy system, or a resource flow operates. Not an improvement on the existing curve. A different curve.
The reason is a bet about competition rather than a bet about science. Breakthrough technologies that solve fundamental scientific challenges typically create their own markets. Incremental improvements face endless competitive battles. We would rather underwrite technological risk than market-timing risk, because technological risk is something our team can actually assess.
One of our portfolio companies, GetSolar, is a clean illustration of the distinction. It did not digitise solar sales. It restructured the entire customer engagement model, compressing deployment timelines from weeks to days and removing friction points that the incumbent model treated as fixed. The technology was not the interesting part. The reconceived process was.
What this means in practice for a founder: if your pitch positions you against three named competitors on a feature comparison, you are describing an incremental improvement, and we are the wrong fund.
Commercial validation, not commercial narrative
Real market traction is the strongest early signal we have that a technology addresses a genuine problem rather than a theoretical one. Paying customers. Signed contracts. Partnerships with something at stake on the other side.
One company in our portfolio secured four million euros in supply agreements before we invested. That is the shape of evidence that moves us — not a letter of intent, not a pilot with a friendly corporate, but a counterparty willing to commit money to a schedule.
This matters more in deep tech than in software, because deep tech pitches are unusually good at sounding validated. A memorandum of understanding with a ministry is not demand. A paid feasibility study is closer. A supply agreement is demand.
Deep Tech Investment Criteria: The Scale-Readiness Thresholds
The question of when a deep-tech innovation is ready to scale is where most founders and most investors talk past each other. I do not think it can be answered on instinct, so we run it against metrics across three dimensions.
Technical Readiness — Deep Tech Investment Criteria in Practice
Technology Readiness Level of at least 7 to 8 — proven in a real operational environment, not a laboratory
Reliability above 99% uptime or the equivalent for the system in question
At least 1,000 hours of operation in real-world conditions
Small-batch manufacturing yields above 95%
That last one is the most frequently missed. If you cannot hold yield at low volume, scale will magnify the problem rather than solve it. Manufacturing scale-up does not fix a process that is marginal in the pilot line.
Market pull
Three to five paying customers beyond pilots, ideally with repeat orders
Clear return-on-investment data validated by those customers themselves — for example a 20% cost saving or a 15% productivity gain that they measured, not that you modelled
A sales pipeline running at two to three times the next year's revenue target
Scalability infrastructure
Margins of at least 40% at ten times current production volume
Dual sourcing for critical components
Capital intensity below 50 cents per dollar of revenue
When those converge — high TRL, positive gross margins at scale, and a demand pipeline already more than double our capacity — that is the point to pour fuel on the fire without burning out the engine. Before that convergence, additional capital tends to buy motion rather than progress.
What we write, and when
Initial cheques range from $300K to $1M, with follow-on capacity extending to $5M for companies demonstrating exceptional progress. We focus on seed through Series A.
That stage choice is deliberate and it addresses a specific gap. Seed to Series A is where a scientific innovation has to prove commercial viability while simultaneously building a business, and it is the stage where many technologies need patient capital and specialised technical assessment that generalist venture funds are not structured to provide. It is also the stage where our capital and operational support can still meaningfully change a company's trajectory. At Series C, we would be along for the ride.
Where we look, and why it is probably not where you are
About 80% of our investments originate from leading research institutions — often in Singapore, Israel, Northern Europe, or leading US accelerators. We then support their expansion into Southeast Asian markets.
This is the whole model, and it exists because of a mismatch. Southeast Asia accounts for a significant share of global emissions growth while undergoing rapid industrialisation and urbanisation affecting 650 million people. Climate technology investment remains disproportionately concentrated in established markets. Solutions are needed in one place and capital is deployed in another.
Two consequences for founders. First, more than 40% of our deals come through trusted, mission-aligned networks rather than open scouting, so a warm introduction genuinely matters more than a cold form. Second, we spend real time on pre-commercial research through the Earth Venture Foundation — grants and sponsorships to academic initiatives, laboratories, and scientific conferences — which gives us sight of technologies 18 to 24 months before commercialization. If you are a PhD student rather than a CEO, that is the door.
What disqualifies you
Every investment must demonstrate a clear pathway to addressing at least two UN Sustainable Development Goals, with quantifiable potential for emissions reduction. Our Investment Committee assesses sustainability implications for every candidate, and we apply strict exclusionary screening against companies that could accelerate fossil fuel usage or environmental degradation.
I want to be honest about the limits of this. Measuring the climate impact of a seed-stage advanced nuclear company or a breakthrough materials startup is genuinely hard — current impact is near zero by construction, and the potential is a reshaped industrial paradigm that conventional metrics do not capture. We assess direct emissions reduction potential, systemic transformation capability, and regional scalability together, and we treat the methodology as unfinished rather than solved.
The part that is harder than the science
The biggest structural problem in funding climate tech today is a mismatch between the urgency of the problem and the pace at which capital is willing to move.
Many investors want short payback cycles and asset-light models. Carbon capture, next-generation energy storage, and industrial decarbonisation involve multi-year R&D, heavy infrastructure, and complex supply chains before meaningful revenue. That does not fit a conventional venture timeframe.
Valuation risk compounds it. Climate impact markets are still forming and policy can move quickly — a change in subsidy or carbon pricing can swing project economics by 20 to 30% overnight. Investors price that uncertainty in, and founders feel it as a discount they did not earn.
Our answer is to blend capital sources rather than pretend one instrument fits the whole journey: venture equity for the innovation phase, project finance for deployment, strategic partnerships for market access. We also make unit economics and policy dependencies explicit from the outset, so risk gets priced rather than avoided.
There is more climate capital available today than at any previous point. The constraint is not the quantity of money. It is matching the right kind of capital to the stage and risk profile of the technology, and that is a design problem the industry has not yet solved.
Design for scale on day one
One last criterion, because it shapes everything after the cheque. We look for technology architected to be globally modular from the beginning — roughly 70 to 80% of the product standardized for consistency worldwide, 20 to 30% left flexible for local adaptation. Engineering, manufacturing, and quality systems then scale without being rebuilt in each market.
The founders who struggle most in our portfolio are not the ones with the hardest science. They are the ones who built for one market and discovered that the second market required a different product.
Build the bridge while people are already walking across it. But design it, from the first drawing, to carry the traffic you expect later.
Editor notes: Tien Nguyen is General Partner at Earth Venture Capital, a deep-tech venture fund investing in breakthrough technologies to make the Earth more efficient. The criteria described here were first set out in an interview with Connect Group, "From Lab to Market — Earth VC's Playbook for Scaling Climate Innovation."
Building something that fits the deep tech investment criteria above? Pitch to us.
Comments