What Does It Really Take to Keep Building When the First Ten Startups Don’t Blow Up?
Soon Kon Chun, CARSOME

Soon Kon Chun’s path into entrepreneurship began in secondary school in Singapore, where a work attachment at a law firm led him to books on business and, eventually, to the word entrepreneurship. That early curiosity carried into the Raffles Entrepreneurship Network, where he took on the presidency and started his first venture straight out of high school, an advertising company focused on restroom spaces. More than ten projects and startups followed, including a B2B marketing agency serving clients such as 3M and an internship platform working with companies like Grab and ByteDance that connected over 20,000 students with opportunities. 

His full-time career has moved from brand management at Procter & Gamble to a startup business development role at Amazon Web Services in ASEAN, and then to CARSOME, which he joined as a Series C company and helped through fundraising during its rapid growth. He describes his approach to leadership as the opposite of micromanagement, and his outlook on the coming years as optimistic about the freedom AI could create.

What first drew you to building within startup environments?

Soon traces his interest back to a school attachment and a single word that stayed with him.

“Back in secondary school, there was a programme that attached students to different companies, and I was placed at a law firm in Singapore. As a student, I could not contribute much beyond proofreading legal documents, so I was often bored and would read articles about entrepreneurship and business online. That was where I first came across the word entrepreneurship. 

Later, in high school, I had to choose extracurricular activities, and one of them was the Raffles Entrepreneurship Network. I remembered the word, applied, and got in, and the rest is history. I eventually took on the presidency, interacted with founders, and that sparked my interest in doing my own startups. Immediately after graduating high school, before university, I started my first venture, an advertising company focused on restrooms, and many more followed from there.”

What has working across many ventures taught you about execution and adaptability?

Having moved from a global corporation into a fast-scaling startup, Soon reflects on operating without a blueprint.

“My first job was in brand management at Procter & Gamble, which required a lot of processes and systems precisely because everything you do impacts millions of people immediately. I then jumped into CARSOME when it was still a Series C startup, which felt like joining a rocket ship. Speed is extremely important in an environment where everything is ambiguous, and you never know what is right or wrong. 

There are no predefined processes, so as an individual you have to consider all the factors and the current situation of the company, then forge the path yourself and create a system that is scalable. That was uncomfortable at the start, because you never know the most efficient way to do something. You do it first, notice the blockers, and work your way around them, and that is where adaptability becomes essential.”

What changes in the Southeast Asian startup ecosystem have had the biggest impact?

Beyond AI, Soon points to the macroeconomic shift that followed the pandemic.

“Aside from AI in the last two to three years, the biggest change has been the macroeconomic conditions after COVID. When I joined CARSOME, my first month saw zero revenue because we could not open our inspection centres. As restrictions eased, we rode a wave of growth, since customers preferred private transport over public transport, which drove huge demand for used cars, and our fundraising went very well. 

From 2021 and 2022 onwards, many similar tech companies in the US were hit by the aftermath of rapid, uncontrolled growth, and investors started to worry less about revenue at all costs and more about profitability as a key metric. That mindset shift changed the whole ecosystem, from cost-cutting measures to a stronger focus on the bottom line. Even now, the conditions are not the most favourable, so we still have to navigate a lack of funding and the impact of AI on the workforce and on businesses themselves.”

What do organisations struggle with most when scaling operationally?

Soon describes how growth can outpace the systems meant to support it.

“The key issue is that systems and processes are not able to catch up with rapid growth. When a company is small, it does not make sense to build too much structure, because that slows the business down. When you scale fast and gain many more colleagues, partners, and customers, sticking with ever-changing processes becomes extremely inefficient. It is not technical debt so much as business debt that you carry forward into your operations, and you become even busier because of these inefficiencies, without the time to clean them up. 

As you scale, it is important to step back and see what can be automated or systematised. At CARSOME, we kept getting asked the same questions about acronyms and business terms, so I created a company glossary and had people add new terms as they encountered them. Little things like that save a great deal of onboarding time, and you can underestimate how much time is wasted through inefficiencies.”

What will the next phase of innovation look like in Southeast Asia?

Soon is bullish on AI, and expects execution to shift decisively towards machines.

“I am very bullish about AI. The next three to five years should be enough for AI to reach a stage where it can do much of the work that white-collar professionals do. I envision founders and business leaders having more time to create ideas, while execution and building are largely taken over by AI. The key task becomes deciding what is the right thing to do, based on the macroeconomic environment and potential competitors. 

Products and solutions are no longer the moat or the unique selling point, so we have to think about network effects or proprietary data that can be scaled as a defence against others, then go all out on execution with AI. Figuring out what to do next is going to be the real work of the next few years.”

What has become most important in your leadership philosophy?

Soon explains why he sees detailed instruction as wasted effort, and trust as an amplifier.

“I am the complete opposite of a micromanager. If I give someone a specific task with detailed instructions, I am wasting my own time creating that guide, and there is little point in having another person simply follow it. My philosophy is to align on the key objective from the start, then let the individual explore how to get there and come up with their own ideas and next steps. 

When they share their thinking, I give feedback based on past experience and context, pointing out what did not work before or where another team is doing something similar. Giving context, rather than specific tasks, is the real role of a leader. By showing trust, that individual gains a greater sense of ownership, and the quality of what they deliver improves. It is a win-win.”

What skills or mindset matter most for people entering the startup world?

Drawing on his time at AWS, Soon returns to a principle he considers non-negotiable.

“One thing that stuck with me from my short time at Amazon Web Services is that customer first is extremely important. It ties back to how leaders will spend the coming years figuring out the right thing to do, which should always be worked backwards from who the customers are, what they want, and what pain points they face. There is no point building something that has no customers or users. 

My biggest piece of advice to students and younger people is to go out and talk to potential customers and understand their pain points deeply. In an AI world, your understanding of your customers becomes a real moat, especially if your solution addresses a deep-rooted problem or a complicated workflow. With AI, small teams can launch great products, so rather than fighting for a huge market, it is better to find a niche segment and serve it very well, which raises the probability of success.”

What do people most underestimate about building and operating startups?

Soon challenges the assumption that early ventures should succeed on the first attempt.

“Most founders’ first few businesses do not do well, including my own. People who have not tried assume that their first or second business needs to succeed, but in reality, doing a startup involves a large element of luck. Success is not guaranteed simply because you did well once, since timing and macroeconomic factors heavily influence the probability. 

Going by sheer quantity is the strategy that many successful founders take, so my advice is not to be discouraged when an early idea does not work out. It is often the fifth or sixth venture, or a later pivot, that actually does well, so you need the endurance to persevere until you get there.”

What is the one question founders and leaders should be asking more often?

Soon frames the defining question around AI’s exponential progress and the future it may create.

“The key question is where our business needs to be in the next three to four years, given that AI is advancing exponentially. I have built SaaS products and solutions, but the latest models can do almost anything, so the real question is what remains for us to build and do. Even a niche idea can be learned and adapted to by AI over time, so we all need to be asking what we are going to do next. 

I am optimistic that we will eventually have a lot of freedom and free time, which is why entertainment interests me so much. I am developing an AI game because I believe entertainment will not be replaced. It is something people will chase after when there is less work to do, and that is the future I envision.”

Join the Conversation: The Ortus Club’s Executive Network

As Soon Kon Chun’s perspective shows, the challenges of building and scaling, from surviving macroeconomic shocks to deciding what remains worth building in an AI-driven world, are rarely resolved in isolation. Founders and operators increasingly rely on candid peer-level dialogue to test assumptions, compare strategies, and refine how they execute.

His emphasis on persistence, adaptability, and customer understanding reflects a broader reality: the most resilient leaders treat volatility as a constant and keep an open mind about where the next opportunity lies. At The Ortus Club, we host curated executive roundtables that bring together senior leaders facing these exact challenges. Step away from the hype and engage in the kind of open, high-value conversations that connect global strategy with practical execution.

FAQs Section

Q: Why do experienced founders treat quantity as a strategy?

A: Because most first ventures do not succeed, seasoned founders start many businesses to raise the probability that one eventually works. Timing and macroeconomic factors play a large role, so persistence across attempts matters more than any single launch.

Q: What is the most important skill in an ambiguous startup environment?

A: Speed. With no predefined processes, leaders must weigh the current situation, forge a scalable path themselves, act quickly, and then adapt as blockers appear rather than waiting for a perfect plan.

Q: How did the post-COVID period change how startups are valued?

A: Investors shifted away from growth at all costs towards profitability as a key metric. This reshaped operations, prompting cost-cutting measures and a stronger focus on the bottom line rather than top-line revenue alone.

Q: What is the biggest operational risk when scaling quickly?

A: Systems and processes failing to keep pace with growth. This creates a form of business debt that compounds through inefficiency, so leaders should regularly step back to identify what can be automated or systematised.

Q: How should founders position themselves in an AI-driven market?

A: By building a moat through network effects, proprietary data, and deep customer understanding, then serving a well-chosen niche segment rather than competing broadly, while using AI to accelerate execution.