The machines were worth $4 million. They were calibrated to tolerances measured in microns. The customers included aerospace MROs, semiconductor equipment makers, medical device manufacturers — names you would recognise. The owner hadn't drawn a salary in eleven months.
Every morning he walked the floor, nodded at his workers, checked the production schedule, and said nothing. Because saying something meant admitting what he already knew. The factory was producing. The business was dying. And he was the only one who knew both things at the same time.
Singapore's precision engineering sector didn't disappear. It moved upmarket.
The labour-intensive, low-margin work went to Vietnam, Malaysia, Johor. What stayed in Singapore — what had to stay in Singapore — was the work that required aerospace certifications, semiconductor-grade tolerances, medical device compliance. Singapore now accounts for roughly 20% of global semiconductor manufacturing equipment production. The aerospace MRO sector has nearly doubled since 2022. These are not struggling industries.
But here is what nobody talks about at the industry conferences. The founders who survived by going upmarket — who made the right strategic call, who invested in the right certifications, who won the right customers — many of them are in serious financial trouble. Not because they got the strategy wrong. Because they got the strategy right and never restructured the financial model to match it.
Pattern one: The receivables trap — worse at the top of the value chain.
The bigger the customer, the longer they take to pay. A listed aerospace MRO. A multinational semiconductor equipment manufacturer. A global medical device company. These are real customers with real purchase orders and real contracts. They are also organisations whose finance teams operate on 90-day payment terms. Sometimes 120.
I have sat across the table from founders with $6 million in outstanding invoices and $180,000 in their operating account. The invoices were real. The customers were legitimate. The work had been done to aerospace standards. But the rent was due. The payroll was due. The raw material supplier wanted payment before the next delivery.
Pattern two: The certification trap — you can't replace what took a decade to build.
To supply aerospace components in Singapore, you need AS9100 certification. To supply semiconductor equipment manufacturers, your processes need to meet specifications that take years of refinement to achieve consistently. Medical device supply chains require ISO 13485 and in some cases FDA registration. These certifications don't transfer. They live in the organisation — and more dangerously, they live in its people.
The senior machinist who has internalised twenty years of rejection patterns. The QC manager who can look at a component and know within seven seconds whether it will pass customer inspection. I have walked factories where the entire quality capability of a $15 million certified business lived inside one person's hands and eyes. And most of them are in their fifties.
Pattern three: Winning the anchor customer that becomes the anchor around your neck.
A major aerospace MRO or semiconductor equipment maker as a customer feels like arrival. What the founder hasn't calculated is what happens when that customer represents 65% of revenue and decides to dual-source. Or brings certain components in-house. Or simply renegotiates because they know he can't afford to walk away.
I have seen founders accept margin compression year after year from a single large customer because the alternative — losing the contract — felt unsurvivable. By the time the customer reduced their order volume, the business had forgotten how to win anyone else. The sales capability had atrophied. The pricing had been suppressed for so long that the team didn't know what the work was worth in the open market.
What a COO sees that the founder often can't.
The founder is too close. He built the certifications. He trained the people. He won the contracts. What I bring into these situations isn't industry expertise. It's distance.
I can look at $6 million in receivables and say: this is not a cash flow problem. This is a contract terms problem, a customer concentration problem, and a working capital structure problem — and all three need to be addressed simultaneously or solving one makes the others worse. I can look at the senior machinist and say: this person needs to be on a structured knowledge transfer programme starting now. Not when he hands in his resignation. I can look at the 65% customer and say: you have eighteen months to reduce this to 35% — or you are not running a business. You are running a dependency dressed up as a business.