Notes from Web Summit 2023 in Lisbon — the week we learned that deep tech is a translation problem before it is an engineering problem.
November 2023BROADMIND
Before Lisbon, our team had spent years in rooms where the word "electrode" needed no explanation. Web Summit is not that room.
In November 2023 we went to Lisbon for the eighth edition of Web Summit, in the deep-tech cohort. It is the largest technology conference in Europe and it is deliberately built to be too big to see all of: 70,236 attendees from 153 countries, a record 2,608 startups, 906 investors, 321 partners and more than 2,100 journalists, all in the Altice Arena and the FIL complex at Parque das Nações.
For a company building non-invasive EEG hardware, that is either the best opportunity of the year or an expensive lesson. It turned out to be both, and the lesson was more useful.
The room is optimised for a different physics
Web Summit's centre of gravity is software. The vocabulary of the floor is ARR, CAC, burn multiple, time-to-value. A SaaS founder can answer "what does it do?" in one sentence and be understood by anyone in earshot.
A BCI company cannot. "It measures electrical activity on the scalp and decodes it into control signals" is accurate and completely useless to a product manager from a fintech. The first two days were, honestly, a crash course in noticing which words we were hiding behind.
The useful ones turned out to be simple:
What does a person do with it? Move a cursor, trigger an action, complete a task in software — without touching anything.
How long does it take to start? Minutes, or it does not ship.
What can I build on it? Raw data, documented, with an SDK. Not a black box.
And the one nobody asks but everybody means: is this real or is it a demo? In a hall where half the booths are prototypes and half are screenshots of prototypes, that question is the entire game.
Deep tech is a category, not an audience
One thing that surprised us was how the "deep tech" label worked in practice. In Lisbon it functioned as a legitimate badge — it told investors this will take longer and cost more than you are calibrated for — but it did not create a shared language. Deep tech at Web Summit meant quantum, photonics, space, advanced materials, biotech and neurotech standing in one aisle, each explaining itself to people from the other five.
That turned out to be useful in an unexpected way. When five founders from five unrelated hardware fields independently describe the same problem — long validation cycles, capital that wants to see revenue in eighteen months, prototypes that cannot be demoed convincingly on a table — you stop interpreting it as your own bad luck and start treating it as a structural feature of the category.
What a generalist conference is actually good for
We came to Lisbon expecting investor conversations. What we got that was worth more:
Compression. Fifteen seconds per person, repeated for four days, forces you to find the shortest true sentence you can say about your product. That sentence did not exist before the trip. It exists now, and it is shorter than what we had in our deck.
Adjacent demand. The most interesting conversations were not with investors. They were with people in enterprise XR, accessibility software, training simulation and industrial monitoring — people with a real problem whose existing solutions (controllers, voice, gesture) break down in specific situations: gloved hands, noisy environments, hands-free requirements. Those conversations did not produce a roadmap that day, and they kept informing it for a year afterwards.
Calibration on our own claims. Conference floors are a good place to be asked to justify a number. If you say "lab-quality," someone will ask which lab, at which sampling rate, with which reference. The correct response is not a bigger adjective — it is a datasheet.
What we could not do there
It is worth being explicit about the limit. A four-day generalist conference cannot validate a scientific claim. Nobody at Web Summit was going to tell us whether our signal chain holds up under motion, or whether our decoding generalises across users. That feedback comes from a different kind of room entirely — and a year later, we went to that room too.
What Web Summit can do is put you in front of the people who decide whether a technology gets adopted, and give you the disconcerting experience of hearing your own pitch out of your mouth for the hundredth time. That is worth the flight.
The short version
The largest tech conference in Europe is a translation test, not a validation test.
Deep tech is a badge, not a shared vocabulary — you still have to explain yourself from scratch.
The most valuable conversations were adjacent to our market, not inside it.
The sentence you need does not exist until you have said it three hundred times.
We left Lisbon with fewer leads than we had expected and a much better sentence. Later, when the Research Kit went to a crowdfunding campaign — 224 backers and $159,616 from people who had never stood at our booth — it became clear that the sentence mattered more than the leads did.
BROADMIND builds non-invasive EEG hardware with unrestricted raw-data access and an open, developer-friendly software path. If you are working on hands-free interaction, get in touch.
Image note: no official Web Summit 2023 photograph of BROADMIND exists in the public record. Supply a team photo from the trip, or run this without a hero image.