Metrics that Matter for Climate Deep Tech at Series A & B

Lessons from Enduring Planet & Friends' April 2026 Community Event

Deep tech doesn't scale like software. But it still gets funded on metrics.

At our latest Enduring Planet & Friends Community Event, "Metrics that Matter for Deep Tech at Series A & B," Dimitry Gershenson (Enduring Planet) led a conversation with Hannah Friedman (Lupine Finance), Supratim Das (Electrified Thermal Solutions), Noah Geeves (Deep Science Ventures), and Laurie Menoud (At One Ventures) on what’s actually being underwritten, from the perspectives of operators, investors and company-builders. This event was sponsored by Wilson Sonsini Goodrich & Rosati.

The takeaway: the metrics problem in deep tech is not about having the wrong numbers. It's about having the wrong framework for which numbers tell a fundable story. Raising at Series A and B in hardware, climate, and industrial is less about hitting benchmarks and more about demonstrating the right proof points at the right time while being able to narrate why they matter.

We structured the conversation around four themes: what metrics actually mean in this context → what unlocks Series A → what makes Series B different → capital efficiency when nothing is linear.

Throughout this post, you'll see italicized callouts marked From the Chat. These capture real questions from founders in the Zoom session, with the panel's live responses. The Enduring Planet & Friends community is built around questions like these. Subscribe to our Luma calendar to bring yours to the next session. 

1. The Wrong Playbook Is Costing Founders Rounds

Most VC frameworks were built for software. Series A means $1M ARR. Series B means repeatable growth. Net revenue retention tells the story. Deep tech founders walk into rooms calibrated for that reality and get evaluated against benchmarks that don't apply to them.

Realized revenue is often the wrong metric entirely for deeptech. A company doing a few million dollars in revenue growing 50% year-over-year might look fine for SaaS. For a hardware company at Series A, it tells investors almost nothing useful. It doesn't prove you can scale to $100M. It doesn't de-risk manufacturing. It proves you found a niche.

What sophisticated investors are actually looking at instead:

• Is the problem you're solving among the top two or three priorities for your customer, not a nice-to-have?
• Are customers willing to pay, and do you know at what price and at what scale?
• Do you have binding proof of demand: not LOIs, not pilots, but offtake agreements or take-or-pay contracts?
• Can you articulate unit economics at scale? (Most companies can't.)
• Do you have people on your team who know what it will take to scale manufacturing?

The distinction that matters: exponential demand before production vs. linear traction after. Companies with zero revenue and $50M in contracted demand close oversubscribed rounds. Companies with $10M ARR struggle at reasonable terms for months. One proved that revenue will scale rapidly once production does. The other proved it can operate, but not that it can scale.

Metrics are also an internal tool, not just an external one. The right framework helps founders build, track, and adjust their path to scale rather than just explain it to investors. Founders who treat metrics purely as fundraising theater tend to be surprised when diligence surfaces gaps they should have seen coming.

2. Series A: What Actually Triggers It

There is no single answer, but there is a consistent structure to how good companies think about it.

Identify your three to four critical path inflection points and treat everything else as noise. Not all milestones are equal. The ones that matter are the ones you genuinely cannot progress without. These typically fall into technical, commercial, and financial categories: proving the core innovation works outside the lab, securing bankable demand commitments, and locking in the capital required to deploy a first unit. Everything else is supporting context.

The trap: founders who build their milestone framework in isolation and optimize for the wrong things for two years. The deep tech field has accumulated hard-won knowledge about what gates scale, but it isn't codified the way software institutional knowledge is. Teams waste significant capital reinventing answers that already exist. Speaking to later-stage investors and founders who've been through it before you need capital from them is not optional. It's how you avoid expensive wrong turns.

On technical readiness, context matters. TRL 5/6 with a pilot online and promising performance data is a reasonable benchmark for many hardware companies at Series A, but modular approaches, high-capex technologies, and novel science pathways each carry different risk profiles. Investors with technical depth can accept more engineering risk if the commercial risk is clearly contained. Most generalist investors want at least one of those off the table before they commit, if not both.

The bar on commercial proof has moved. Investors now expect multi-million dollar signed customer agreements beyond pilot stage. Not just letters of interest, but commitments with specificity on pricing, volume, and timeline. Detailed LOIs with real terms can be meaningful; vague ones are nearly worthless. The signal investors are looking for: is this customer real and serious, or are they just willing to evaluate if you produce something they like later at a price they haven't agreed to?

From the Chat: The chicken-and-egg of binding commitments. Founders pushed back on the offtake bar: how do you secure binding demand before you can scale production? Detailed LOIs with pricing formulas, volume schedules, and clear customer deployment commitment can bridge this gap when they're investor-quality. One founder shared their Series B is being underwritten on $1B in LOIs from investment-grade buyers, with product still 12 to 18 months from field deployment. Sophisticated investors engage with that structure when the underlying commitments are real, especially when they can call the customer themselves and confirm.

For industries that don't structurally support long-term offtakes, chemicals or cement for example, the answer is to make whatever commitment you can get as detailed and investor-quality as possible. Sophisticated investors can work with structural constraints. What they can't work with is vagueness dressed up as progress.

From the Chat: Government and federal customers. For founders selling primarily into federal or state government, the question came up repeatedly: these customers structurally can't make the kinds of forward purchase commitments VCs want to see. The workaround that builds credibility is demonstrated progress on SBIRs, cooperative agreements, and other structured government engagement, treated as a milestone progression rather than as offtake. 

Financial infrastructure matters as much as the metrics themselves. Investors who ask hard questions in diligence expect auditable models with clear assumptions, scenario analysis, and a financial operator who can stand behind them. Founders who have to reconstruct data mid-process signal that the discipline behind the numbers isn't there. That costs deals, sometimes quietly, in the form of term adjustments or investor confidence that erodes without ever being named.

3. Series B: The Transition Nobody Talks About

Series B in deep tech is structurally harder than Series A. Investors want to see scale. But deep tech scaling is non-linear. You can't hire more salespeople and watch revenue compound. The growth story doesn't fit a standard model, and founders who try to tell it that way lose credibility with investors who've seen enough hardware companies to know the difference.

The shift from technology development to project development is real, and it shows up in the metrics. At Series B, the capabilities that matter change. Procurement professionals. Supply chain depth. Manufacturing and engineering expertise on the team. A pipeline of maturing projects with enough similarity to suggest repeatability. Founders who haven't made this transition yet will show it in how they talk about scale: long on technology confidence, short on project execution conviction.

Manufacturing strategy is one of the highest-stakes decisions a hardware company makes. Outsourcing to contract manufacturers is more capital-efficient in most scenarios and leverages established expertise. But for complex equipment operating at extreme temperatures, pressures, or chemical conditions, ceding control over critical manufacturing processes carries real operational risk. The right answer depends on what the technology actually requires, not on what looks cleanest to investors.

The four factors to weigh explicitly: speed, cost, diversification, and vision. The right manufacturing approach is the one that's honest about your specific business model and end state, not the one that fits a slide.

From the Chat: Building deployment muscle in-house vs. relying on EPCs. Founders post-seed asked how early investors expect to see independent deployment and engineering teams. The answer depends on your end-state: if you intend to be the long-term builder-owner-operator of facilities, keep deployment expertise close. If you don't, paying the premium for external EPCs makes sense, but smaller companies often get the B-team within large engineering firms, and lacking an exceptional internal project manager will blow out timelines either way. Working toward a reputable EPC partner is itself a credibility signal. 

Revenue velocity matters more than revenue level. Lumpy revenue, one deployment then a gap then the next, is the reality for many hardware companies. The question is whether deployments are becoming faster and more predictable, and whether the second unit is cheaper and easier than the first. That trajectory is what investors are underwriting, not a smooth growth curve that doesn't reflect how physical infrastructure actually scales.

Series B investors often ask questions that founders aren't prepared for around team structure and financial engineering. Capital stack management. Contingency planning. How to layer in project finance or debt as the business matures. These aren't CFO-only concerns. They reflect how the company intends to finance the next phase of growth without continuing to rely purely on dilutive equity. Founders who haven't thought through this tend to give vague answers that signal the work hasn't been done.

4. Capital Efficiency When Nothing Is Linear

Capital efficiency is a metric that means something very different when you're building physical infrastructure. Borrowing the SaaS definition makes most hardware companies look worse than they are.

Burn multiple and runway tell different stories depending on what mode you're in. R&D burn, pilot-stage burn, and deployment-stage burn have different risk profiles and different expected outcomes. Framing all three the same way invites the wrong comparison. The more useful narrative: what milestone is this burn funding, and what does the cost structure look like once you're past it?

From the Chat: Licensing and capex-light models. Founders flagged a recurring frustration: VCs encourage licensing, then pass because licensing isn't seen as backable. The nuance: capex-light doesn't have to mean you can't produce, it can mean you externalize production to contract manufacturers while retaining IP and margin. But for most deep tech, you still need to build and operate something in the early stages to demonstrate you have anything worth licensing in the first place. 

Thoughtful investors can help plug gaps in capital stack management. The companies that use capital most effectively tend to be working with financiers who provide structured guidance, not just checks, and who help them think through how to sequence equity, debt, grants, and project finance across the development lifecycle. Honest and open conversations about risks and mitigants tend to produce better partnerships and better outcomes than fear-driven positioning.

From the Chat: Policy and regulatory risk. With perceived stability of the US policy environment eroded, this came up repeatedly. The framing from the panel: policy can be a powerful tailwind, but it cannot be the foundation of a fundable business. The companies that raise well show economic viability on their own merits, with policy treated as accelerator rather than dependency. 

Investors should be asking about team composition more than they do. Across the conversation, one theme came up repeatedly: the metric that most reliably predicts whether a deep tech company will scale is whether the team has the right people for the next phase of the business, not just the current one. Technical differentiation matters less than founders think. Demonstrated customer behavior and the organizational capability to execute on it matters more.

A few things investors ask about frequently that often aren't the most informative signal:

• Gross margins in isolation, without context on where you are in the cost curve
• Revenue level as a proxy for de-risking, when what matters is velocity and contractual commitment
• Burn as a negative, without reference to what the burn is building

What they should ask about more: manufacturing expertise on the team, how founders are thinking about the transition to project development, and how the capital structure will evolve as the company moves from equity-funded pilots to project-financed commercial deployments.

Final Takeaways

Deep tech fundraising is a translation problem as much as a metrics problem. The underlying proof points, a technology that works, demand that's real, a team that can execute, are the same things investors have always cared about. The challenge is expressing them in a framework that VC rooms can evaluate, not based on errant benchmarks. Use metrics and milestones to tell a story that demonstrates your critical path is clear, your demand is locked in, and your team can execute the next phase of scale.