Deep tech companies raise from two different kinds of money and usually tell both parties the same story. Early capital is buying a mission it can commit to for a decade. Later capital is buying proof that somebody will pay. The materials that serve one do not serve the other, and building the wrong set is an expensive way to lose a year.
Two kinds of money
An investor I listened to recently described the first kind bluntly. You have to talk someone who is a bottomless pit of capital into believing your long-term vision, he said, because it is almost a non-logical thing for them to give you money. The fundamental science is going to consume an enormous amount just to get you to the start line.
That is not a criticism of those investors. It is an accurate description of what they are doing. When the payback sits fifteen years out and depends on physics that has not been demonstrated at scale, no model produces a defensible number. What remains is a judgement about whether the mission is worth committing to and whether this team is the one to do it.
The second kind is the opposite. By then the technology risk is largely retired and the question is whether the output has buyers, which is a question with an answer. The same investor’s description of that end of the market was that they care a lot less about your technology and a lot more about who is going to buy it.
Same company, different decade, completely different conversation.
What belief capital is actually buying
It is buying a reason to care that survives fifteen years of technical setbacks.
The Boring Company is the standard illustration, and it holds up. The physical activity is digging a hole in the ground, which nobody would fund enthusiastically. The story was an automated high speed transit network with a line drawn all the way to tunnelling on Mars. Identical activity, and a completely different set of people willing to work on it and put money behind it.
Deep tech founders resist this, and the resistance is honourable rather than stupid. Academic training rewards precision and punishes overstatement, so describing your work in terms of where it could lead feels like the beginning of dishonesty. What gets missed is that the alternative is also a choice. Describing a battery sensor as four millivolts of measurement accuracy is precise and tells the listener nothing about why it matters, which is that electric vehicles catch fire and this is how you find out before they do.
The material that serves this end of the market is about people and destination. A founder who can explain, on camera, why they gave up an academic career for this. Photography of the actual facility and the actual team. A film that shows the problem in the world rather than the equipment in the room. It works because the investor is making a judgement about people and conviction, and neither of those survives a slide deck.
What evidence capital is actually buying
Names, mostly.
Which companies have evaluated this. Who has paid for a pilot. What the procurement conversation looked like and how far it got. Whether the people who would have to buy this at volume have heard of you, and what they said.
The material that serves this is completely different in character. Case studies with real detail. Film of the technology working in a customer’s environment rather than in your own laboratory, because those are not the same proof. A website that answers the questions a procurement team asks, which are about supply, specification, support and risk, not about the underlying science. Campaign infrastructure that produces a documented pipeline rather than general awareness.
I have written separately about what that evidence is worth at the point of a raise, in why your next raise depends on your buyer list.
Where this goes wrong
Two failures, and the second is more common than the first.
The first is telling a belief story to evidence money. A founder walks into a conversation with people who fund facilities and talks about transforming an industry, and the room goes quiet, because that is not the question they asked. They wanted to know who has signed.
The second is leaving the belief material in place too long. A company builds a good vision-led site during its early years, it works, and then three years later the same site is still leading with the mission while the company is trying to sell to procurement teams who need to know about supply guarantees. Nobody decides to do this. The site just stops being updated in step with the company.
The tell is usually that the homepage describes a future and the sales conversations are about a present. When those two drift apart, the site starts costing you deals rather than winning them.
Real limitations
Since this is a piece about telling a more ambitious story, it needs a clear limit, and the limit is tense.
Describe what exists in the present tense and what you intend in the future tense, and never let a picture imply a stage you have not reached. A rendering of a plant, labelled as a rendering, is a legitimate and useful thing to show. The same image presented so a viewer assumes it is a photograph of an operating facility is not, and in a small industry you will be found out by someone who knows exactly how far along you are.
The same applies to film. Showing a prototype that is genuinely held together with tape, shot properly and framed as a prototype, is more persuasive than a polished animation of a product that does not exist, because the tape is evidence that something real happened. Buyers in this market are technical. They can tell the difference, and being caught blurring it costs more than the additional ambition ever earns.
We have turned down a shoot on this basis, where the story being requested was two years ahead of the equipment.
Which one are you raising?
Two questions usually settle it.
Can anyone build a defensible financial model of your return, or does it rest on a judgement about whether the science will work? If it rests on the judgement, you are raising belief capital and your materials should be about mission, team and destination.
And is the main open question in your investor conversations whether the technology works, or whether anyone will buy it? When that question changes, and it changes fairly suddenly, your materials need to change with it, and most companies notice about a year late.
Plenty of companies sit between the two, usually around a Series A, where the round is half funded on conviction and half on early commercial signal. Those need both, which is a sequencing problem rather than a contradiction. Film the founder and the mission once, because that material has a long life. Build the evidence layer continuously, because it decays and needs replacing.
What this means for what you build
If you are raising on belief, spend on people and place. Founder interview film, real photography of the facility and the team, and a site that leads with the problem in the world. That work is cheaper than most founders assume and lasts for years. The approach we take to filming people who would rather not be filmed is in how we film founders who hate cameras.
If you are raising on evidence, spend on proof and reach. Customer-environment film, case studies, a site built for a procurement team, and campaigns that produce a named pipeline. That is mostly growth work rather than brand work.
If you do not know which you are raising, do not book either yet. Go and find out, and the fastest way is usually customer discovery.
A 20 minute call will get you a straight read on which of these your company needs. The Albotherm project ran the website first and the photography and film after it, which is the usual order when the commercial story is settled and the evidence is still being built.
Frequently asked questions
What is the difference between belief capital and evidence capital?
Belief capital funds work whose commercial return is too distant to model, so the investor is committing to a mission and a team over a decade. Evidence capital funds work whose return can be calculated, so the investor is assessing whether demand exists. They ask different questions and respond to different material.
How do you tell an investor story without overclaiming?
Keep the tense honest. Describe what exists today in the present tense and what you intend in the future tense, and never let an image imply a stage you have not reached. A rendering labelled as a rendering is fine. The same rendering presented as a photograph of a working plant is not.
Should a deep tech company make a film before it has a product?
It can, if the film is about the problem and the destination rather than a product demonstration. That is often the most useful film a pre-product company can have, because it gives investors, recruits and early partners something to react to. The mistake is filming a product story when there is no product.
What changes in your materials between an early round and a later one?
The centre of gravity moves from why this matters to who is buying it. Vision material stays but stops leading, and named customers, pilot results and procurement-ready detail move to the front. Most companies leave the early material in place too long.
Can the same film work for investors and customers?
Partly. The founder and problem sections usually work for both audiences. The proof sections do not, because an investor wants to see a market and a buyer wants to see their own operation. Filming once and cutting differently for each audience is cheaper than two productions and is how we normally plan a shoot.