Building The Anti-Fragile Startup In 2025

Forbes

20-07-2025

Most startups are built for one thing: rapid scale. Raise capital, hire fast, grow revenue, repeat. This playbook has a better shot at working in stable markets with predictable growth trajectories. But when uncertainty hits – economic downturns, supply chain disruptions, AI transformation – these growth-optimized organizations often shatter.
A different breed of start-ups is emerging, one that builds what philosopher Nassim Taleb calls "anti-fragile" organizations. Unlike resilient systems that merely withstand stress, anti-fragile systems actually improve under pressure. They use volatility as fuel rather than viewing it as an obstacle to overcome.
How start-ups operate when they are anti-fragile resembles how nature looks amidst storms.
After interviewing executives who've scaled companies through multiple crises, five core principles separate anti-fragile organizations from their fragile counterparts.
1. Make Vulnerability Part of How You Operate
Traditional startup culture worships the myth of the infallible founder. But leaders building anti-fragile organizations do something counterintuitive: they systematically embed vulnerability into their operational DNA.
Antonio Silveira, CTO of Attentive, learned this principle while navigating economic turbulence across multiple companies. "When I make mistakes, I address them in my all-hands. I say, 'Hey, I heard the feedback. This was my intent. This is what I learned. This is what we're going to change.' You need to emulate that so others feel like they can keep giving me feedback."
This isn't performative – it's strategic infrastructure. When leaders model fallibility during stable periods, they create organizational antibodies against the blame culture that paralyzes decision-making during actual emergencies. Teams that regularly practice acknowledging mistakes develop what scientists call "error recovery systems" – the ability to learn from failure faster than competitors can avoid it.
Curtis Anderson, CEO of Nursa, took this to its logical conclusion by requiring childhood photos for internal profiles. "You could scroll through the Slack directory and it was just everybody as a third grader," he explains. "There's something about that visual that makes everyone totally approachable."
When people aren't spending mental energy protecting their image, they redirect that bandwidth toward solving problems.
2. Treat Recovery Like Training, Not Time Off
Most startups treat recovery as what happens after burnout. Anti-fragile organizations flip this equation, building recovery into their operating system as a performance enhancer rather than a performance penalty.
Lorraine Buhannic, Chief People Officer at Headway, knows this firsthand as the mental health platform has scaled from zero to 700 employees in four years. "We have what we call our Olympic performance standard," she explains. "We expect people to do the best work of their careers here, and we want that to feel motivating and galvanizing and not overwhelming and lead to burnout."
The Olympic analogy is precise: elite athletes understand that recovery isn't the absence of training—it's training for the nervous system. "One explicit point in that principle is around the importance of recovery. Olympic athletes need to have recovery in order to perform at the highest level," Buhannic notes.
At Headway, this translates into structural expectations: flexible PTO policies, therapy sessions openly blocked on calendars, and what Buhannic calls "transparent mental health infrastructure." The sophistication lies in framing recovery as performance optimization rather than accommodation for weakness.
Organizations that normalize recovery as operational necessity build sustainable competitive advantages over those that optimize for short-term intensity.
3. Build for Learning Speed, Not Just Performance
Traditional performance management was designed for industrial environments where roles were stable and best practices were known. Anti-fragile organizations require performance systems optimized for learning velocity rather than measurement accuracy.
Doug Dennerline, CEO of Betterworks, learned this lesson painfully during his tenure managing 6,000 people at Cisco. "We used to do bell curve ratings where you were forced to have a top 15% and a middle 75%," he recalls. "They've updated it since, but I still remember how horrific those processes were. You're telling 75% of the population that you're just mediocre?"
His solution wasn't to abandon measurement but to reconstruct it around adaptive capacity. "We've tried to create lightweight, in-the-moment points in which managers can give relevant feedback to employees. Schedule feedback along the way, one-on-ones on a weekly basis."
Frequency transforms feedback from judgment into coaching. When feedback becomes continuous rather than episodic, it shifts from performance evaluation to performance development.
Buhannic at Headway operationalized this through what she calls "contextual competence" – embedding management development within the specific cultural framework of the organization rather than teaching generic leadership skills.
In volatile environments, learning velocity matters more than any specific knowledge or skill set.
4. Stop Trying to Resolve Every Contradiction
Most strategic frameworks seek consistency and clarity. Anti-fragile leaders have learned to embrace paradox – holding apparently contradictory truths simultaneously and using that tension as competitive fuel.
Rajat Bhageria, founder of Chef Robotics, embodies this principle. His path from venture capital to robotics entrepreneurship taught him a crucial lesson: "As an entrepreneur, you have to be irrational. You just keep going even though you're beat up literally every single day. Whereas as an investor, you have to think about what are all the risks and be very risk averse."
Rather than choosing one mindset, Bhageria learned to toggle between them strategically. "You have to have the irrationality of a founder, but at the same time you should try to de-risk the business. If you find some fundamental reasons why the business is not going to work, then you should be real about this."
This meta-cognitive ability – thinking about thinking – becomes crucial during uncertainty. While competitors get paralyzed trying to resolve contradictions, anti-fragile leaders use them as navigation tools.
Lucia Huang, co-founder of Osmind, applies this to technology adoption. Rather than either rejecting or blindly embracing AI, she holds the contradiction: "We have to remember that a lot of our end users are actually quite skeptical about AI and it has a lot of potential to do harm in our space too. So we're trying to approach it with a really cautious, clinician-first approach."
While others are stuck in either/or thinking, leaders who master both/and thinking can navigate complexity with nuance.
5. Make Everyone an Experimenter, Not Just a User
The AI revolution is creating a fundamental split: organizations treating technology as tools to be deployed versus those developing technology as cultural capabilities to be evolved. The second approach builds sustainable advantages.
Huang at Osmind demonstrates the sophisticated approach: ""We've done an AI sprint internally to upskill and up-level our team. We supplied budgets for everyone to experiment with AI. This internal experimentation allows us to thoroughly test and refine AI capabilities before integrating them into our platform, ensuring we deliver proven, reliable tools rather than experimental features."
This isn't just individual development—it's building organizational learning systems. When every team member becomes an experimenter rather than just a user, the organization develops distributed intelligence about technological possibilities and limitations.
Dennerline at BetterWorks takes this further: "I have been pushing this pretty hard. I let people experiment with it. I pay people bonuses to come up with ways that produce increases in productivity. We had a woman on our India team implement a thing that uses AI to do QA that used to take us four weeks, but now takes us four hours."
The crucial insight comes from recognizing that technology adoption is ultimately about human psychology. As Buhannic observes: "People's expectations of work has changed a lot. Managers need to be more than just a person directing work. They're really setting culture."
Successful technology integration requires maintaining human agency while augmenting human capability. It's about creating hybrid intelligence that's more powerful than either humans or machines alone.
Building for What's Coming
Anti-fragile organizations don't just survive disruption; they use disruption as raw material for competitive advantage. They understand something their fragile counterparts miss: in a world of accelerating change, the ability to improve under stress becomes more valuable than the ability to avoid stress altogether.
As Anderson puts it: "The future is for the faithful. That's not for everybody. But then you show incremental progress, and you help people understand that each of these points on the line is directionally moving the way that we need it to."
This isn't blind optimism. It's earned confidence that comes from building systems designed to get stronger when the world gets stranger.
I write about the intersection of AI and performance management for Forbes. I'm the founder of Mandala, an AI Coaching Platform for Managers.