What a deep tech video agency does differently
Why a generalist video agency struggles with deep tech, what an engineer changes in a film, and the questions worth asking before you hire anyone.
- A deep tech video agency is one where somebody on the film side can read your documentation without a translator sitting next to them.
- The generalist failure is not bad craft. It is a film that looks expensive and describes a category rather than a company.
- An engineer changes the film before production, not during. The expensive corrections are the ones made after rendering.
- Ask for the technical review stage by name. An agency that has no answer has never had its facts checked by a client's engineers.
A deep tech video agency is a studio where somebody working on the film can read your spec sheet, your architecture diagram or your paper without a translator in the room. That is the whole distinction. Everything else, the craft, the 3D, the sound design, is available from several hundred studios who will do it well.
Why does a generalist video agency struggle with deep tech?
Because the brief degrades before anyone writes a word. A generalist studio takes a briefing call, does not follow the technical middle of it, and writes to the part they understood. What comes back is competent and looks expensive, and it describes your category rather than your company.
You can usually see it in the first fifteen seconds. The film opens on a claim that any competitor could make: the industry is changing, the old way is broken, here is the future. That opening exists because it is the only ground the writer felt safe on.
What does an engineer actually change in a film?
The order of operations, mostly. Which sounds procedural and is worth more than it sounds.
- The script gets specific earlier. Someone who understands the mechanism can write about the mechanism, so the differentiator survives into the voiceover instead of being smoothed into a benefit statement.
- The wrong technique gets rejected at the brief. 3D is not automatically right. It is right when the object does something a diagram cannot show.
- Factual errors get caught before rendering. This is the expensive one, and it is covered below.
What is technical review, and why does it decide the budget?
Technical review is a stage where your engineers check the film for factual accuracy before anything is animated. It is the single stage that separates a deep tech film from a SaaS film, and skipping it is the most costly decision available on the project.
The reason is arithmetic. A wrong claim caught in the script costs a sentence. The same claim caught after a 3D sequence has been modelled, lit, animated and graded costs the sequence. Deep tech films fail on facts far more often than on craft, and facts are cheap to fix in week one.
If an agency cannot describe this stage without being prompted, they have never had a client's engineers mark up their work.
Should the technique be decided in the brief?
No, and deciding it there is the most common expensive mistake. Choosing 3D at the brief stage is choosing the answer before anyone has asked the question.
The technique follows the script. Once you know what has to be understood, the form is usually obvious: an invisible process wants 3D, a workflow change wants motion graphics, a claim about trust wants a person on camera. Picking the form first means writing backwards from a look.
How long should a deep tech film be?
Shorter than the brief usually assumes, and driven by what has to be understood rather than by how much there is to say.
| Length | What it can carry |
|---|---|
| Under a minute | One idea, one proof. A launch, a product moment, a conference loop. |
| Around two minutes | A mechanism explained, with the context that makes it matter. |
| Beyond three | A specialist audience that has already chosen to pay attention. |
Our own films sit mostly around the two minute mark, which is where a mechanism can be explained without the middle sagging.
What should you ask a video agency before you hire them?
Four questions, and the answers are more diagnostic than the reel:
- Who on your side will read our documentation? A named person, not a process.
- When do our engineers see it? If the answer is "at review", ask whether that is before or after rendering.
- How do you choose the technique? If 3D is proposed before the script exists, that is a look being sold, not an argument.
- What happens when we disagree about a claim? Your engineers are usually right about the fact and usually wrong about the emphasis. A studio that has done this before will say so.
When is a film the wrong investment?
When the thing you need is not comprehension. A film is expensive and slow relative to a page, and it earns that when the mechanism is genuinely hard to picture: something invisible, something at a scale nobody can observe, something whose value only appears in motion.
If your problem is that nobody can find you, the film is the wrong first purchase. If your problem is that people find you, read the page, and still cannot tell what your technology does, that is precisely what a film is for.
Does an engineer-led studio actually change the day to day?
It removes the glossary call. That is the plainest way to say it.
The brief stops degrading between what you explained and what got written down, technical review gets shorter because fewer things are wrong by the time your team sees it, and the render is right about what the object does rather than only how it looks. None of that is a claim about taste. It is a claim about how much of your meaning survives the trip to the finished film.
Written by Mejo Kuriachan. More in the blog, the glossary and the FAQ.