RESEARCH
All research45 funded deep tech websites, read one by one.
The crawlable basics are mostly fine. The layer that decides whether a stranger can act on you is broken on 96 percent.
of 45 funded deep tech sites gave a visitor no named person to email.
OUR AUDIT, SEPTEMBER 2026 · 45 SITESLIMIT · SIX SITES COULD NOT BE CONFIRMED. UNKNOWN, NOT FAILURES.The question.
A funded deep tech company has one job for its website before revenue: let a stranger who is already interested act. We read 45 of them, one by one, to see whether a stranger could.
How we did it.
Computed from the audit sheet built on 14 and 15 September 2026, recomputed and corrected on 18 September. 52 sites were read; 45 carry the full audit columns, so every percentage is out of 45. Median site size is six pages.
What it means.
The technical hygiene these companies could be expected to get wrong, they mostly get right. The thing they get wrong is the one that decides whether a stranger who is already interested can act.
The four tests, ranked.
SHARE OF 45 THAT FAILEDTHE ORDER IS THE FINDING: THE BASICS A CRAWLER NEEDS ARE MOSTLY FINE. THE ONE A PERSON NEEDS IS NOT.
Who, exactly.
NAMES NO HUMAN BEING, BY STAGEMore than half of the youngest companies.
A third, the same as the whole sample.
It resolves by Series A. Five companies, so read it as a direction, not a law.
How to reach a person.
CONTACT ROUTE, OUT OF 45THE SIX UNCONFIRMED ARE UNKNOWN, NOT FAILURES. WE FOUND EXACTLY ONE DIRECT EMAIL. WE DO NOT CLAIM ONLY ONE EXISTS.
What to do with it.
THREE THINGS TO CHECK ON YOUR OWN SITEName a person.
Put at least one named human being on the site. A third of the sites we read had none.
Give them a direct email.
A general mailbox is not a person. We found one working direct email for a named person in 45 sites.
Serve the words without JavaScript.
Five of 45 sent no body copy to a crawler or a language model.
ONE QUESTION
Could a stranger email a named person at your company in under a minute?
Yes or no. If it is no, that is the first fix.
EVERY FIGURE HAS A SOURCE
Sources and limits.
- SOURCEOur own audit of funded deep tech company websites, read one by one on 14 and 15 September 2026 and recomputed on 18 September.
- SAMPLE52 sites read, 45 kept with the full audit. Every percentage is out of 45.
- LIMITSSix contact routes could not be confirmed because the page never loaded. They are unknown, not failures. The cuts by stage rest on 5 to 27 companies each, so read them as direction.
- ALSO IN THE DATAPlatforms, where identified: WordPress and Elementor 13, Webflow 11, Next.js 4, Wix 2, Lovable 2, one-off builders 2. Eleven could not be identified.
Send your site. We name the first fix.
One line about where you are going is enough. The reply comes back in writing. No call needed.
Or book a 30 minute call
