How to choose a deep tech branding agency
What to actually evaluate, the tradeoffs between a generalist and a specialist, the questions worth asking, and the answers that should end a conversation.
- Ask who will actually do the work. The gap between the pitch team and the delivery team is the single most reliable predictor of how a deep tech engagement goes.
- A generalist with a strong process beats a specialist without one. A specialist matters when the failure mode is misdescribing the technology, which in deep tech it usually is.
- Ask what happens if the work can never be published. A studio whose growth depends on publishing has a quiet incentive to push you toward disclosure.
- The portfolio to ask for is the one where the product has no visual interface. Anyone can make a robot look good.
We are one of the agencies you might be evaluating, so read this accordingly. It is still the honest version, because the criteria below are the ones that decide whether the engagement works, and we lose some of them.
What should we evaluate first?
Who does the work.
The most reliable predictor of how a deep tech engagement goes is the distance between the people in the pitch and the people who deliver. Ask directly: will the person who understood our technology in this meeting be writing the copy, or does it pass to a team we have not met?
This matters more in deep tech than elsewhere because the understanding is the expensive part. A studio that has to re-explain your mechanism internally will produce work that is one translation removed from the truth, and you will spend the engagement correcting drift.
Generalist or specialist?
Depends on which failure mode you are more afraid of.
A generalist with genuinely strong process will get the brand system, the craft and the discipline right, and will need you to supply the technical judgment. That works if you have someone internally with the time and taste to be a real reviewer.
A specialist should already speak the domain, which removes the glossary call and, more importantly, removes the risk of a page that is elegant and subtly wrong. The tradeoff is a smaller portfolio and less exposure to how other categories solve things.
The honest test: is your biggest risk that the work will be badly made, or that it will be well made and inaccurate? In deep tech it is usually the second, which is the argument for a specialist. If it is the first, hire the better studio regardless of category.
What portfolio should we ask to see?
The one where the product has no visual interface.
Anyone can make a robot look impressive. The hard case is a company whose advantage is a protocol, a material, a signalling scheme or a model, where there is nothing photogenic to point a camera at. Ask for that work specifically. If a studio only has visually generous categories in its portfolio, you are its experiment.
Ask what the work was trying to achieve, not just how it looked. A studio that can only talk about the visual decisions did not make the strategic ones.
What questions actually separate agencies?
Five, and the answers are more revealing than the deck.
Who reviews the technical claims, and when? The right answer involves your engineers, before anything is designed. If the answer is a proofread at the end, factual corrections will arrive when they are most expensive.
What happens if the work can never be published? Some deep tech work is permanently under NDA. A studio whose growth depends on publishing its work has an incentive, usually unspoken, to push you toward disclosure you are not ready for.
Will you name us as a client, and when? Plenty of studios will sign an NDA about your technology and treat your logo as theirs to display the day the site launches.
What do you do when we disagree about a technical simplification? The answer you want is that they lose that argument to your engineers on facts and win it on structure. A studio that always defers produces accurate, unreadable work. One that never defers produces the opposite.
What is the timeline and what is fixed? Fixed scope and a fixed date put the risk of a slow week on the studio. A day rate makes every hour of confusion billable, which quietly aligns their interest against yours.
What is a reasonable evaluation process?
Three conversations, not a formal RFP.
An RFP for brand work mostly measures who is best at writing RFP responses. Instead: an initial call where they try to understand the technology, a second where they show relevant work and you interrogate the decisions, and a third with the people who would actually deliver.
Between the second and third, send a real piece of your material, a spec sheet or a technical page, and ask what they would change and why. The quality of that answer, given for free in an hour, tells you more than any case study. Studios that will not do it are telling you something too.
How much should the shortlist be about price?
Less than most procurement processes assume, and not zero.
The variance between credible studios on a comparable scope is smaller than the variance in outcome. The expensive failure is not overpaying by twenty per cent, it is a brand that misdescribes the product and has to be redone in eighteen months, or one that is never finished because scope was never fixed.
What price should tell you is whether the scope is understood. A quote that arrives without anyone asking what the film has to survive, or who reviews claims, is a quote for a different project.
When is the answer none of them?
When the position is not settled and nobody internally can settle it.
If the shortlist itself is settled and the open question is what kind of engagement to buy, we have written separately on choosing between a diagnosis-first project and a design subscription.
No agency can decide which category you are in if the founders disagree and the disagreement is unresolved. Good studios will surface that and some will decline. If three separate agencies have told you the same uncomfortable thing about your positioning, that is data, not a sales technique.
Equally, if the problem is that nobody has heard of you, brand strategy is not the first purchase. If the problem is that people find you, read the page, and still cannot tell what you do, that is exactly what this is for.
Written by Mejo Kuriachan. More in the blog, the glossary and the FAQ.