Before you fund the factory, fund the evidence a factory needs

The hardest early funding problem is often not a shortage of capital. It is an oversized question. A charging depot, greenhouse, mini-grid cluster or pilot plant can prove something specific: customers will pay, equipment will perform, collections work, or a plant can produce to specification. Equity, grants, strategic customers and lenders want different things from that proof. The founders who match each capital source to its risk create a path to the next cheque. Those who treat a factory and a company as one financing problem can discover that an announcement is not cash.

Britishvolt had the shape of a national success story before it had the cash to build one.

The UK battery company planned a gigafactory. It had been offered a £100 million Automotive Transformation Fund grant. Then it sought an early £30 million draw. The money was not available because the grant was payable in instalments after milestones, including private-investment commitments, had been met. Britishvolt had not met them. No grant money was paid. In January 2023, the company entered administration.

That is not a story about a grant disappearing. It is a story about treating a future factory and a young company as one enormous financing question.

Here is the payoff before we go any further: a founder facing that question needs a way to shrink it. The Proof Asset Test at the end of this article does exactly that. It asks what one physical asset can prove next, who can afford to fund that uncertainty, and what must happen before promised money is actually spendable.

MIT Energy Initiative found that venture firms put more than $25 billion into cleantech startups between 2006 and 2011 and lost over half of it. Hardware, materials, chemicals and industrial processes performed especially badly because they required patient capital, long development periods and tolerance for commodity-market economics. Venture capital was often being asked to do a job it was not built to do.

The obvious answer is to raise more equity.

The useful answer is stranger: raise for less.

Fund the proof. Not the whole company.

The milestone Britishvolt did not have

Most infrastructure founders face a familiar pressure. They need equipment, a site, construction, people and customers at roughly the same time. So the pitch becomes one large request: fund our technology, facility and expansion plan.

Imagine you are an investor hearing that request. You are being asked to judge whether the technology works, whether the asset will be built on time, whether customers will buy the output and whether the company can operate it. Any one of those may be manageable. Together, they are a thicket.

Britishvolt’s grant terms made the problem painfully visible. Public support existed, but it was conditional. It could not bridge the period before the conditions were satisfied. Subsidies can exist without becoming spendable cash.

The central distinction is not debt versus equity. It is company risk versus asset risk.

Company risk includes a young team, unproven technology, changing plans and overruns. Asset risk concerns a defined thing: a greenhouse, two buses, a production plant, a charging depot or a group of mini-grids. Once that asset has customers, contracts and operating data, a different kind of capital provider can assess it.

That sounds neat on paper. The real cases are messier.

Two buses changed the question

In Nairobi, BasiGo did not begin by asking bus operators to take the full financial gamble on electric transport.

The Kenyan company used roughly $1 million of pre-seed capital and a $4.3 million seed round in February 2022 to import two buses, install a charging depot and run a pilot. It retained ownership of the batteries and charged operators per kilometre under its Pay-As-You-Drive model. The expensive battery and charging equipment were separated from the operator’s purchase of the bus.

That was the clever part. The pilot did not merely prove that an electric bus could drive through Nairobi. It tested whether route operators could use a different buying model.

TechCrunch reported the seed round. GET.invest, which supports sustainable investment preparation, later described how BasiGo aligned its revenue streams, legal capital structure and lead debt contracts before approaching specialist e-mobility financiers. This is where most people stop looking. The buses mattered, but so did the paperwork around them.

According to GET.invest, BII, the UK development-finance institution, provided $5 million of early-stage debt after comparable commercial debt proved unavailable. In October 2024, BasiGo raised $24 million of Series A equity, unlocking $10 million from DFC and a further $7.5 million from BII. DFC’s seven-year facility formed part of a $35 million project intended to expand from 19 buses towards 2,075 by 2028.

Here is the twist: equity was not replaced by debt. Equity made the debt possible.

An equity investor buys ownership. Its return depends on the company becoming more valuable, so it can tolerate uncertainty while the model is being proved. A lender expects repayment. Its return is interest, not an unlimited share of the upside. That is why it wants credible revenue, asset performance, contracts and protection against early losses.

BasiGo used two buses and a Nairobi charging depot to test a different ownership model for electric transport. Photo: Bahnfrend / Wikimedia Commons, CC BY-SA 4.0.

The verbal demand that vanished

Husk Power Systems learnt a harsher version of the same lesson in India.

Its founders financed their first two rural power plants with $95,000 in business-plan prize money. Shell Foundation then supplied grants of $125,000 in 2008 and $270,000 in 2009. India’s Ministry of New and Renewable Energy provided subsidies. Husk pre-sold about $800 to $900 of electricity per site.

Then some of that demand vanished before customers were connected.

At first glance, this looks like a sales problem. It was more consequential. A mini-grid is not financeable merely because it generates electricity. It has to collect cash, maintain equipment and operate across sites without collapsing under management strain.

Husk raced to more than 80 biomass plants. Shell Foundation reported in January 2018 that shortages of management talent, smart metering, collection systems and long-term debt constrained the portfolio. The company restructured towards a less capital-intensive model and eventually solar-hybrid mini-grids.

The proof asset in this case was not simply a working power plant. It was an operating site that showed whether households actually paid, meters worked and equipment could be maintained. That is a much tougher test.

A grant behaves differently again. It does not seek an ownership return like equity, or scheduled repayment like debt. It can fund an experiment or public outcome when the unknown is too early for a lender. But it offers no commercial validation on its own. A grant can pay for learning. It cannot make reluctant customers pay their bills.

Husk shows why technical generation is not enough: collections, maintenance and operating systems decide whether mini-grids can scale. Photo: Kneerajp / Wikimedia Commons, CC BY-SA 4.0.

Porsche bought the learning

On a windy site in Chile, Haru Oni offered a third route.

The e-fuel pilot brought together Porsche, Germany’s economy ministry and industrial partners including Siemens Energy, Enel, ExxonMobil, Gasco and Chilean state oil company ENAP. Porsche committed an initial €20 million and became the pilot fuel’s principal user. Germany’s economy ministry supplied approximately €8.2 million.

Porsche’s published project information described a roughly $78 million facility on 3.7 hectares, using a 3.4 MW wind turbine and a 1.2 MW electrolyser. Its 2021 announcement projected growth from 130,000 litres annually to 55 million litres in 2024 and 550 million litres in 2026.

Porsche did not behave like a normal lender. It was a strategic customer. Its return was not simply interest, and not merely a future ownership gain. It gained access to fuel, technical learning and a demonstration relevant to its own business.

That distinction is easy to miss because strategic capital often arrives wearing the clothes of a corporate announcement. A strategic customer may be willing to finance learning that a conventional lender cannot underwrite. But a projected scale-up is still only a projection. The original announcement does not prove that commercial-scale economics will follow.

Haru Oni shows how a strategic customer can fund demonstration value before commercial-scale economics are proven. Photo: Ildottoreverde / Wikimedia Commons, CC BY-SA 4.0.

When a plant becomes a different investment

Twelve, the US company turning carbon dioxide into jet fuel and chemicals, made the separation unusually explicit at AirPlant One.

In September 2024, Twelve announced $645 million: $200 million of Series C corporate equity, up to $400 million of project equity led by TPG Rise Climate, and $45 million of credit facilities. A separate $25 million project-finance construction loan supplied non-dilutive working capital. The project also had customer validation through a 14-year IAG agreement covering 785,000 tonnes of fuel.

AirPlant One began commercial operations in June 2026. On 9 September 2026, Nomura and Endurance Capital refinanced the construction facility with an up-to-$45 million credit line after the plant was producing specification-compliant fuel and naphtha, according to Twelve.

The important event was not just the large funding announcement. It was the moment construction uncertainty became operating evidence. Only then could the construction facility be refinanced.

Put the cases side by side. BasiGo’s core risk was whether fleet operators would adopt a new payment model; two buses and a depot were the proof asset; seed equity came first; the milestone was financeable operating and contractual evidence. Husk’s risk was whether rural demand became collected cash; its proof asset was a functioning mini-grid site; prizes and grants funded learning; the milestone was reliable operations, not merely generation. Haru Oni’s risk was technical integration and future fuel demand; its proof asset was the pilot; strategic and public money paid for demonstration; the milestone was learning for industrial partners. Twelve’s risk moved from construction to production; its proof asset was AirPlant One; corporate equity, project equity and credit each took different exposure; the milestone was specification-compliant output.

None of these is a generic ‘funding journey’. Each is a deliberate handover of one risk to the party best able to carry it.

Twelve separated corporate equity, project equity and credit, then refinanced after the plant produced specification-compliant output. Photo: Dmitry Kharitonov / Pexels, Pexels licence (free commercial use).

The paperwork that became an asset

Pure Harvest Smart Farms, an agricultural-technology company in the United Arab Emirates, began with a less glamorous expenditure.

Its founders invested more than $1 million to acquire a 3.3-hectare site and pay for engineering, feasibility, hydrogeology, legal and structuring work. That made the first greenhouse shovel ready. Government-backed capital, technology partners and specialist angels then financed a $4.5 million seed round for the commercial-scale proof facility.

After demonstrating year-round production, Pure Harvest obtained a $50 million, three-year sukuk with embedded warrants in 2021, alongside $10 million of equity, to construct further facilities.

What surprised us was how often the missing asset is not the equipment. It is the evidence pack that allows someone else to understand the equipment, site and contracts. Feasibility work may feel like overhead. In infrastructure, it is often the thing that converts a hopeful plan into something capital can examine.

Pure Harvest used founder capital for the diligence that made its first commercial greenhouse financeable. Photo: Lyn Ong / Pexels, Pexels licence (free commercial use).

When the proof asset is still the wrong answer

Staged finance is not magic. Sometimes the smallest credible proof asset is still too expensive, too slow or too uncertain to attract an appropriate first funder.

That happens when a pilot cannot produce a meaningful commercial answer, when the cost of the demonstration is nearly as large as the eventual build, or when no customer can make a useful commitment before construction. A pilot may prove chemistry but not the economics of producing at scale. It may prove a machine works but not that it can be operated, maintained or paid for.

Britishvolt is the warning at the far end of that problem. A giant factory cannot easily be broken into a cheap, self-contained proof point if the money needed before the proof is already vast and conditional. In those cases, founders need to reduce the technical or site uncertainty further, find a strategic participant with a reason to pay for learning, or accept that the project is not yet ready for staged finance.

The lesson is not that every big project can be made small. It is that nobody benefits from pretending it already has the evidence of a smaller one.

What founders should do before the next pitch

Start with one physical proof asset. Name it precisely. Two buses and a depot. One greenhouse. A defined cluster of mini-grids. A pilot plant.

Then answer the uncomfortable question: what commercial fact will it prove? Not that the equipment exists. What will it prove about payment, demand, uptime, maintenance, collections or contracted use?

Prepare the site, engineering, feasibility, legal structure and customer evidence before you ask a lender to believe in the asset. Treat every grant condition as a funding risk. Do not raise on empty: a signed-looking commitment is not spendable until its conditions, timing and draw process are clear.

Finally, do not send every capital provider the same story. A grant body needs to understand the experiment and public outcome. A strategic customer needs to see what it gains by participating. An equity investor needs the company opportunity and the uncertainty being removed. A lender needs to see how it will be repaid.

What experienced investors see first

Investors are not necessarily sceptical of big visions. They are sceptical of undifferentiated risk.

The first question is often not, ‘How large is the market?’ It is, ‘What exactly does my money fund, and what happens if the milestone slips?’ Equity investors know they may lose everything, so they look for ownership upside large enough to justify that exposure. Grant providers look for learning or policy impact. Strategic funders look for access, capabilities or supply. Lenders look for protection against not being repaid.

Put the cases side by side and a pattern appears that none of the announcements quite says aloud: the strongest early infrastructure round is often designed around who can afford to be wrong about one particular thing.

Porsche could fund learning about e-fuels. Shell Foundation could fund mini-grid experimentation. Equity investors could absorb BasiGo’s early commercial uncertainty. Lenders could arrive once BasiGo and Twelve had created more protection beneath them.

GI Network’s view: Infrastructure founders should not ask, ‘Who funds companies like mine?’ Ask, ‘Who is structurally able to carry this next uncertainty?’ The answer changes as the asset moves from idea, to construction, to operating evidence.

GI Network would isolate the next proof asset and test whether its cash flows, customer commitments, site work and legal structure can withstand investor scrutiny. We would map which costs belong in company equity, grant-backed experimentation, strategic participation or later credit; identify conditions that could delay drawdown; and rehearse the objections an investment committee will raise before outreach begins.

The Proof Asset Test

Before seeking capital, run seven questions in order:

1. 1.The thing: What exact asset are you funding now?
2. 2.The proof: What commercial fact will it establish beyond technical performance?
3. 3.The evidence: Which operating measures will show that proof?
4. 4.The user: Who will buy, use, operate or benefit from the output, and what can they commit today?
5. 5.The preparation: Are the site, engineering, feasibility and legal foundations ready for diligence?
6. 6.The risk match: Which party can carry this uncertainty: equity, grant, strategic capital or lender?
7. 7.The release: What conditions must be met before each promised sum is actually available?

If those answers produce a smaller ask, that is not a retreat. It is progress. You are no longer asking someone to finance a future empire on faith. You are asking them to fund the next piece of evidence.