top of page
negative_horizontal.png

What we saw at Startup Summit 2026

4 hours ago
5 min read

Design for Scale spent three days at Startup Summit 2026 in Florianópolis. Notes on what the stages were saying, a neuroscientist's counterpoint, a trade floor doing real business, and a question about the MVP that nobody answered the same way.



Ten stages and a lot more around them

Startup Summit 2026 ran from August 26 to 28 at CentroSul. The core of the programming sat on ten stages organized by track, with a second layer of content spaces, activations and pitch areas running in parallel. Close to 300 speakers over three days. We went in with a long list, covered a lot of it and still missed plenty.


Where we ended up says something about how we read the event: the Founder Stage, the AI track and its younger sibling on MicroSaaS and vibe coding, the marketing and sales room, and detours through people, M&A and retail. Some sessions were deep and technical. Others were built to inspire. The programming trusted the audience to want both, and that trust felt right.


The scene that repeated most, on stage and off it, was a founder describing a working prototype built in a weekend and then asking what came next. On the other side of the table, investors described pipelines full of products that all looked capable, and the difficulty of telling them apart came up more than once.



What the whole program was saying

Scan the full program and the pattern is clear before any single session starts. AI has stopped being a track and become a qualifier. It is attached to agriculture, energy, health, security, retail, leadership, sales and hiring, and "in the age of AI" now precedes almost every other topic. Inside that, the center of gravity has moved from generating to delegating. Agents were the most repeated concept after AI itself, and the framing of the founder's job shifted with them, from building to orchestrating. Running against all of it, a steady undercurrent of skepticism: a meaningful share of the sessions were built around the gap between hype and practice, and asked what actually changes when the technology reaches the business.


The business side was just as consistent. Ecosystem was the noun of the event, applied to sectors, to cities and to survival. Scale appeared as a destination and efficiency as its condition. Exits and multiples were discussed with unusual candor. And the origin story, from zero to a number, was the format founders showed up for.


Brand made it to the main stage twice, once as the invisible asset behind the companies that grow and once as positioning, the work of becoming the first choice.



What the rooms we chose had in common

Reading the titles of the sessions we attended side by side, a few patterns jump out, and they tell the story better than any single talk.


The first is the ultimatum. Adapt or die. Ecosystem or death. Translate or be cut. Three different stages, three different speakers, one grammar. The message to founders and to marketing leaders alike was that the window for doing things the old way has closed, and that the cost of staying put is now stated in the title.


The second is growth as something that needs to be sustained. Reputation that sustains growth. The partnership between HR and the CEO that sustains growth. Leading the growth of your company. The verb never appears alone. Growth is assumed; the question on every stage was what holds it up once it starts.


The third is the decision. The architecture decision that defines your business. HR at the decision table. Selling, operating and deciding with more intelligence. Less budget, more traction: the science behind the channels that scale. The event was less interested in what is possible than in how you choose between options that all look possible.



Build, buy or plug in, and then what

The AI track opened with the most practical question in the building: build, buy or API, and which of the three defines your business. The follow-ups were about which type of AI is producing the most value today, beyond text, and which techniques will define the next generation of agents. On the main stage, a panel with people from Replit and Notion presented vibe coding as the new stack from idea to launch. On the Founder Stage, investors from Redpoint and Headline discussed the markets AI is unlocking.


Heard back to back, these sessions were made mostly of open questions. The ease of creating a product has generated a set of needs most teams have not named yet. Positioning, distribution, brand, the ability to be recognized in a category that fills faster than anyone can map it. The tools solved building. They created the problem of being seen.



Brand: lever or luxury?

The marketing and sales room asked the question directly, in a panel on reputation as the foundation of growth. An hour earlier the same stage had hosted a talk on translating marketing into the language of the board so the budget survives the next cut, and before that, a session on the science behind the channels that actually bring scale on a smaller budget.


Read together, the three sessions describe the same tension. Brand is being asked to prove itself in a spreadsheet at the exact moment when product differentiation is getting harder and reputation is doing more of the work. Our answer to the panel's question is the reason Design for Scale exists, and the rest of the event kept supplying evidence for it.



A counterpoint from neuroscience

On the second day, Miguel Nicolelis took the main stage with a talk titled AI: neither intelligent nor artificial. Nicolelis is the neuroscientist who pioneered brain-machine interfaces at Duke and led the Walk Again Project, the exoskeleton controlled by brain activity that opened the 2014 World Cup. He has spent years arguing that the human brain is not computable and that current systems are pattern machines trained on our own output, with no path to the thing we call intelligence.


It was the most radical version of the skepticism running through the program, delivered by someone from outside the ecosystem with no stake in its bets. In a building full of people betting their next three years on AI, it was a sharp and deliberate contrast. Some of the audience pushed back. Most of it listened. Either way, it was the talk people were still discussing at dinner, and a reminder that the event had room for disagreement, which is rarer than it should be.



The business floor was on fire

We used the Summit mainly to learn, but the commercial side deserves its own paragraph. More than 11,000 people, around 3,000 startups and 150 investment funds passed through the venue. Over 30 international delegations ran structured business rounds with Brazilian startups. The organizers' final tally put investment intentions generated during the three days at close to R$ 400 million.


On the floor, that translated into packed stands, pitches running in parallel and meetings in every available corner. The exhibitor area worked as a marketplace, with companies at Traction and Scale using their booth as a brand statement as much as a sales point. The ones that stood out had done that work in advance: a visual identity you could recognize from across the hall, printed and digital materials that belonged to the same system, motion on the screens instead of a looping slide deck. Design was the difference between a stand people walked past and a stand people walked into.



Is the MVP dead?

That question came up, in those exact words, more than once. If a working product is a few prompts away, what does building one still prove? Several speakers offered replacements. None of them agreed with each other.


We left with a view on that, and it involves brand design and the assets a company ships before its product. It is the subject of our next article.




Design for Scale

Creative partner for startups & scale-ups

 
 

Ready to scale with
clarity and impact?

Design For Scale is a sister brand of Clint.

bottom of page