Most fundraising advice was written for generic software startups. Our new guide, collating input from ten firms that back frontier founders early, is for everyone else.
In deep tech fundraising, the hardest thing to prove to an investor often isn’t the merit of the idea, but that other investors may also be convinced by it.
In the early days of Valar Atomics, plenty of investors passed for exactly that reason. As founder Isaiah Taylor later put it, they passed because they didn’t believe the company could raise the scale of capital required.
Ramana Nanda and Matthew Rhodes-Kropf call this financing risk; the danger that a company with sound science dies because it’s too difficult to raise the next round.
As silly as it sounds in a supposedly contrarian discipline, investors might pass on a deal they otherwise like because they fear too many other firms will disagree.
Julie Fredrickson of Chaotic Capital wrote the first cheque anyway, backing a long shot because of her belief in Isaiah. In June 2026, Valar’s Ward 250 reactor reached criticality in Utah. According to the Department of Energy, it was the first DOE-authorised reactor built outside a national laboratory.
Chaotic Capital is one of ten firms whose partners shaped Frontier Capital, our fundraising guide for founders building on the frontier of technology.
Why most fundraising advice fails deep tech founders
Fundraising advice reflects the companies venture capital funds most regularly, and for two decades that meant software — mostly following a dull B2B SaaS playbook.
Michael Ewens, working with Nanda and Rhodes-Kropf, traced part of the shift to the launch of Amazon Web Services in 2006. Once startups could rent computing power rather than buy it, the cost of creating a new software startup collapsed.
Investors adapted with what the authors call a “spray and pray” approach: smaller first cheques into more companies, less governance, and a readiness to walk away after the first round. Boom or bust financing, rather than patient capital.
That works well enough for a company that can test an idea for a few thousand dollars. It doesn’t work at all for a company that needs to build a reactor, a pilot plant or a launch a clinical trial before it can get anywhere near commercialisation.
VC’s favourite growth benchmark is built on this reality. “T2D3”, the shorthand Neeraj Agrawal of Battery Ventures coined in 2015, means tripling revenue twice and then doubling it three times; roughly $2M to $144M of annual revenue in five years.
Many frontier companies won’t have meaningful revenue in five years. Even when they do, it will never grow with such predictability.
Unfortunately, this means most fundraising advice online is either useless or actively harmful to founders in frontier categories.
How to find your first believers
The investors most likely to back you early are the contrarians; the ones whose model depends on disagreeing with the market as their source of alpha.
They are unfortunately scarce. Carta’s analysis of more than 900 US deep tech rounds in 2025 found them more dilutive than software rounds at the early stages. This is the expected result when there are a limited number of buyers in the market.
So, the search matters as much as the pitch. Look for firms that have backed similarly ambitious ideas just as early, and skip investors who don’t accept the problem you’re solving. It’s effectively impossible to convert someone who isn’t willing to believe.
When the ten firms in the guide describe what they look for, with their own thesis and approach to investment decisions, the common thread is distinctly not pedigree. It is evidence of commitment to a problem, and resourcefulness in attacking it.
How much to raise, and at what price
Venture capital is the most expensive money a frontier company will raise, and it should be spent incredibly carefully. Think of it as ignition fuel: its job is to secure the evidence that makes the company financeable by cheaper, larger and more specialised capital.
And by the way — that pool is enormous. For example, researchers studying how to finance fusion energy noted that in 2018 the global bond market was more than 400 times the size of that year’s venture investment.
So, price discipline matters as much as the amount. Adam Besvinick of Looking Glass Capital warns that if you raise $5M at a $25M valuation, you have “a target on your back before you’ve even written a single line of code”.
A valuation set with software-like expectations will become a hurdle that a frontier company may need years to clear. Which is not to say they wont exceed those targets in future, but the journey there isn’t likely to be a smooth curve.
Too much money too early, as we’ve argued before, can do as much damage (and potentially more) as too little. Constraint is the mother of innovation.
That discipline matters more now that deep tech is back in fashion. Stefano Bernardi’s October letter to Unruly Capital’s investors describes more capital chasing the same deals, at rising prices.
Momentum-driven capital tends to evaporate as quickly as it appears.
How long a deep tech raise really takes
Decisions might come quickly, if you’re working with good firms who respect your time, but conviction usually doesn’t (and nor should it).
At firms like 1517 Fund, partners have often known a founder for months or years before they invest — although they’re still often the first investor. The investment decision itself is where that conviction comes to a head, not where it begins.
Even then, the raise itself is often slow. In a survey of 885 venture capitalists, the average deal took 83 days to close, including 118 hours of due diligence. The time between rounds has also extended, now closer to two years.
You’ll have to budget for both. The guide suggests starting with at least 12–18 months of runway, and allowing 6–12 months from first conversation to money in the bank.
Then choose investors who will still pick up the phone to help you out in year four. As Arian Ghashghai of Earthling VC puts it, “Venture is a long, brutal war of attrition.”
The ten firms behind the guide
The guide profiles all ten firms involved, in their own words where possible:
1517 Fund, Also Capital, Boost VC, Cantos, Chaotic Capital, Earthling VC, Long Journey Ventures, Looking Glass Capital, True Ventures and Unruly Capital.
Readers of The Odin Times will recognise most of these names. From 1517 Fund’s renegade techno-patronage to Boost VC’s remarkable debut fund, built by backing ideas from science fiction, Earthling’s explicit focus on frontier technology, Looking Glass Capital’s focus on finding value where it is overlooked, and Also Capital’s mastery of origination, and Chaotic Capital’s ability to underwrite uncertainty.
None of this is new to venture capital. In fact, it’s where venture began. The industry’s first great success was a bet on hardware: in 1957, American Research and Development put $70,000 into Digital Equipment Corporation.
Somewhere along the way, people decided software was easier. The industry got lost in financial engineering, ARR multiples, and the mirage of predictable returns.
Indeed, most investors will tell you that hard tech is hard to fund, and they’re right.
That’s precisely why it’s valuable.
Fortunately, you only need the valuable handful who disagree.
All your first believers, one line on your cap table: Founder Rollup, from Odin



