Deep Tech Investor Communications Between Milestones
Australia's newest unicorns build physical things. Their story problem is the gap between milestones
Australian startups raised $1.8 billion in the first quarter of 2026, up 63 percent on the same quarter a year earlier, and the strongest opening to a year since 2022. That was the headline.
Here is the number underneath it. The top 10 deals captured 59 percent of all capital raised, and the top 20 captured 79 percent. By funding concentration, it was the most lopsided quarter in over seven years. Deal activity below $5 million hit its lowest quarterly level since 2020.
So two markets ran at once. A small group of companies raised category-defining rounds, and a much larger group found conditions considerably tighter than the aggregate suggested.
What changed about the small group is the interesting part. All three companies that reached unicorn status in the quarter build physical things: Gilmour Space, a rocket manufacturer valued at A$1.5 billion after a $217 million Series E, Advanced Navigation, an AI-enabled navigation and autonomous systems company valued at A$1.5 billion after a $158 million Series C, and Neara, a critical infrastructure modelling platform valued at A$1.1 billion after a $90 million Series D.
Third Hemisphere is an Australian communications agency working with climate technology, deep technology, and finance companies across Asia-Pacific. Its investor communications work sits on exactly the problem this shift creates, because a company building in atoms has a story timeline that behaves nothing like a software company's.
What is the deep tech narrative gap?
The deep tech narrative gap is the period between a company's technical milestones during which it has real progress to report and no conventional news to report it with. Software companies close that gap with shipped features, customer counts, and revenue growth. A hardware or science-led company can spend 18 months on a single validation step that produces no announcement at all.
The capital timelines make the gap structural rather than incidental. A typical Australian venture fund has seven to 10 years to return capital to its investors, while deep tech ventures often need five to seven years just to reach commercialisation. That leaves almost no room for the growth phase a fund needs, which is a large part of why generalist investors have historically passed.
Where the capital actually went
The sector figures show appetite has changed.
Hardware, robotics, and sensors attracted $303 million across 17 deals in the quarter. Space and defence took $229 million across just three deals, an average of $76 million each. Cybersecurity raised $125 million, led by UpGuard's $105 million Series C, and AI infrastructure took $100 million through Firmus.
Investor sentiment matches the flow. On net excitement, vertical business software led at plus 48 percent, with hardware, robotics, and sensors close behind at plus 45 percent, healthtech at plus 27 percent, and AI models and data infrastructure at plus 23 percent. At the other end, consumer brands sat at minus 53 percent and crypto at minus 48 percent.
Five years ago the atoms categories barely registered in Australian venture portfolios. In the first quarter of 2026 they accounted for the plurality of capital deployed.
Government has moved in the same direction. The NSW Government's $20 million Emerging Technologies Commercialisation Fund offers grants of $500,000 to $2 million to companies at Technology Readiness Levels 3 to 7, sitting inside a broader $79.2 million state innovation commitment. It targets what the sector calls the valley of death, where a technology has early validation and still lacks the capital to become an investable business. Full applications were due in July 2026, with successful applicants expected to be notified in October.
Why concentration changes the communications job
When 20 deals take 79 percent of the capital, being in the visible set becomes a practical concern rather than a vanity one.
Nobody should claim that visibility causes a large round. Technology, team, market, and timing do that. But investor attention is finite, specialist deep tech investors in Australia are few, and the companies those investors have been reading about for two years enter a process differently from the companies they meet cold.
That is the honest version of the argument. Communications does not manufacture a Series B. It determines whether the people who could fund one already understand what the company does before the deck arrives.
For a company with a five-year commercialisation path, that understanding has to be built during the years when there is no funding announcement to hang it on.
What to publish between milestones
Five kinds of material fill the gap without inventing news.
Validation steps, stated plainly. A completed test, a certification achieved, a materials result replicated, a pilot moved from one site to two. Each with the method, the conditions, and what it does not yet prove. Investors read qualifications as competence.
The technical explainer nobody else has written. Deep tech categories are usually badly explained in public. A company that writes the clearest available explanation of its category becomes the reference point for journalists, analysts, and the generalist investors doing early reading.
Named technical people. The engineers and scientists holding the claim are usually the most persuasive people in the company and the least visible. Bringing them forward, with clear boundaries on what is confirmed and what is still under test, builds credibility that another executive byline cannot.
Customer and partner progress at whatever scale it exists. A single utility running a trial is a real proof point when described accurately. Overstated as a partnership, it becomes a liability at diligence.
A position on the policy and procurement conditions the company depends on. Grid rules, defence procurement, critical minerals policy, or standards development. A company with a considered public position on the conditions affecting its market reads as commercially literate.
None of that requires a funding round, a product launch, or a claim the company cannot support.
The specific trap for hardware founders
Long timelines create pressure to fill silence with adjectives. It is the most common failure in the category, and it is the most expensive.
A company that has described itself in superlatives for three years, with no third-party validation attached, has trained its market to discount everything it says. When the genuine milestone arrives, the language has no headroom left. Investors who have watched the pattern read the real result as more of the same.
The alternative is unglamorous and works better. Small, specific, verifiable claims, published consistently, with the limitations stated. Each one is checkable. Together they build a record that a diligence process can follow, and they leave the strong language available for the moment it is earned.
There is a second trap in the other direction. Companies whose science is genuinely difficult sometimes retreat into technical language that only their peers can parse, then wonder why generalist coverage never arrives. A journalist at the Australian Financial Review, a state government adviser, and a superannuation fund analyst all need a version they can use. Writing that version is a discipline rather than a dilution.
Why the audience for this material is wider than investors
Founders often treat investor communications as a separate activity aimed at a small group. In deep technology it rarely works that way, because the same material has to serve four audiences with different needs at once.
Investors want a legible path to commercialisation, with the risks named. Customers, who in these categories are often utilities, government agencies, or industrial operators, want evidence that the technology performs in conditions resembling their own. Government bodies assessing grant applications want technology readiness stated precisely, since the NSW fund's own eligibility turns on sitting between Technology Readiness Levels 3 and 7 and applications outside that band are routinely ruled out. And engineers deciding whether to join want to see technical seriousness.
One well-written validation update can serve all four. A vague announcement serves none of them, and a company producing several of those a year has spent its credibility without buying anything.
The practical consequence is that the publishing cadence should be built once, for the most demanding reader, then distributed. In these categories the most demanding reader is usually the technical buyer rather than the investor, because they will test the claim against their own site conditions.
Two questions for the next board meeting
What have we published in the past six months that an investor could verify? Count only material with a method, a date, a number, or a third party attached. If the answer is nothing, the company has been silent to the market for half a year regardless of how much marketing activity took place.
Could a generalist investor explain what we do after five minutes on our website? Specialist deep tech investors in Australia are few, so most early reading is done by people without domain expertise. A clear category explanation in plain language decides whether they keep reading or move on.
The takeaway
Capital is moving toward Australian companies that build physical things, and it is concentrating in a small number of them. For founders with five-year commercialisation paths, the work is publishing verifiable progress during the years between milestones, so the investors who could fund the next round already understand the company when the process opens.
Third Hemisphere works with deep technology and climate companies on exactly this cadence. Jeremy Liddle (LinkedIn), the agency's Managing Director, has invested in over 25 technology and climate startups, which shapes how the agency reads what an investor is assessing. To review what your company has put in front of the market this year, book a consultation or browse Third Hemisphere's insights.