What deep tech branding needs at seed, Series A and Series B
What the brand and the website actually have to do at each stage, what is premature, and why the same work is right at one round and wasted at another.
- At seed the job is one legible sentence and a page that does not look abandoned. Full identity work is usually premature.
- At Series A the buyer arrives. The site stops being a fundraising artefact and starts having to survive a technical evaluation.
- At Series B the problem is usually incoherence: four surfaces telling four versions of the story after three years of accretion.
- The most expensive mistake at every stage is the same one, building identity on a position that is not settled.
The same brand work is correct at one round and wasted at another. This is what changes, and what is premature.
For where the category as a whole stands this year, read the state of deep tech branding. This post is about what changes from one funding round to the next.
What does a seed-stage company actually need?
One sentence that a stranger can repeat, and a page that does not look abandoned.
At seed the company is usually pre-product or pre-revenue, and the audience is almost entirely investors and early hires. Neither is reading your site for a specification. They are checking whether you can explain what you do and whether the company looks like it exists.
That is a much smaller job than a brand system. A clear positioning line, a page that loads, an honest description of the mechanism, a deck. What is genuinely premature at seed is a full identity system with a brand book, because the position is still moving and everything built on it will be rebuilt.
The exception is naming. If the name is wrong, fix it now. A name gets more expensive to change every month it exists.
Why does Series A change the job?
Because a buyer arrives, and buyers evaluate differently from investors.
An investor is assessing whether the category is real and whether you can win it. A technical buyer is checking whether the thing works, integrates and is supportable. The first audience tolerates ambition; the second is looking for reasons to disqualify you.
That is when the website stops being a fundraising artefact. It needs a product page that survives someone qualified reading it closely, specifications presented so a specialist can find them, and enough evidence that the claim is checkable. Most Series A deep tech sites are still written for the previous audience, which is why the pattern of a strong round followed by a slow sales quarter is so common.
Series A is also where the nine to sixteen week engagement makes most sense: positioning, narrative, identity and the site, in that order, once.
What breaks at Series B?
Coherence, almost always.
By Series B there have been three years of accretion. The deck says one thing, the website another, the booth panel a third, and the product pages were written by whoever had time. Nobody decided to become incoherent; it happened one deadline at a time.
The work at this stage is rarely a rebrand in the sense of a new logo. It is usually a refresh: settle the position again now that the company actually knows what it sells, then make the surfaces agree. Companies that jump straight to a visual redesign here fix the symptom and keep the disease.
The other Series B pattern is a category move. A component supplier becoming a platform company, or a tools company becoming an infrastructure company, is a positioning problem first and a design problem a long way second.
When is a full rebrand actually justified?
Three situations, and none of them is boredom.
The company has moved category, so the old brand describes a business it no longer runs. The name is wrong, blocked or inherited from a lab or parent it has separated from. Or the brand actively misdescribes the product, which is the one case where doing nothing costs more than doing something.
Everything else is a refresh, and a refresh is a smaller, faster and more honest piece of work.
What is premature at every stage?
Building identity on a position that is not settled.
This is the most expensive mistake available and it does not care what round you are at. Changing the category you claim while identity is applied, 3D is modelled and a build is underway costs the identity, the 3D and the build. Changing it in week 3 costs a conversation.
The order is not administrative. Positioning, then narrative, then identity, then the site. Companies that do it in a different order pay for it twice.
Does the deck change by stage?
Yes, and mostly in what it has to survive.
A seed deck is carrying an argument about a category and a team. A Series A deck has to survive a technical diligence call afterwards, which means the mechanism slides get read by someone who could build it. By Series B the deck is competing against companies with revenue, so the burden shifts to evidence: retention, deployments, cycle time, whatever your category counts.
A deck that wins the meeting but loses the diligence is a Series A failure mode specifically, and it is common enough to design against.
What should we do first, at any stage?
Find out how long it takes a qualified stranger to understand what you do.
Show your page to someone technical who has never seen the company, and time how long before they can say the sentence back. If it is under twenty seconds, your problem is somewhere else and you should spend the money there. If it is not, that gap is costing you meetings you never hear about, and it will cost more at every subsequent round.
Written by Mejo Kuriachan. More in the blog, the glossary and the FAQ.