The industrial economy is mostly SMEs. The transformation advice is mostly written for multinationals. It is time the 99% had a blueprint of their own.
For more than four decades, I have walked factory floors across Singapore and ASEAN. The firms I remember best are not the flagship plants with glass-walled control rooms. They are the 40-person stamping shops, the family-run food processors, and the injection moulders who supply the giants, keep the region’s economy moving, and almost never appear in a keynote.
They are the overwhelming majority. Across OECD economies, small and medium enterprises make up roughly 99% of firms and generate somewhere between half and two-thirds of value added. In Singapore, SMEs account for about 99% of enterprises and around seven in ten jobs. And yet, for a generation, the advice these firms receive about technology has been written for someone else.
This series, and the book behind it, starts with a simple claim: the 99% aren’t behind. They have been measured against the wrong benchmark and sold the wrong plan.
Open almost any smart-factory reference architecture, and you will find the same assumptions: a greenfield site, uniform modern equipment, a dedicated data-science team and an enterprise budget. The case studies celebrate lighthouse plants with hundreds of engineers.
For an SME, that advice is not merely unhelpful. It is quietly misleading because it sets a benchmark that makes every gap look like a deficiency to close with spend. The right benchmark for a 60-person factory is not a 6,000-person one. It is the best possible version of a 60-person factory.
When a smaller firm copies a multinational’s architecture, it imports the multinational’s costs: licences priced for scale, integration projects measured in person-years, and a dependence on consultants for every routine change. It does not import the scale, specialist teams, and balance sheet that make those costs worthwhile. I call this enterprise cost without enterprise advantage, and I have seen it drain more SME modernisation budgets than any other single mistake.
If technology were the constraint, the SMEs that bought the most would grow fastest. That is rarely what happens.
The barriers that keep most manufacturers small are organisational, and they reinforce one another:
Buying a platform doesn’t solve any of these. All of them are solved, one step at a time, by making decisions explicit, shareable and better.
The 99% Factory blueprint begins somewhere unexpected. Not with AI, not with a platform, not with a roadmap of forty initiatives. It begins with a single question:
Which recurring decision, if we made it better and faster, would change a customer outcome we are paid for?
Consider Meridian Stamping, one of the composite cases that runs through the book: a 60-person precision stamping firm supplying a Tier-1 customer. In one quarter, it faces three pressures at once. The customer demands lot-level traceability with every shipment. Input costs compress an already thin margin. And its most experienced setter, the man who knows by ear when a tool is about to drift, announces his retirement.
The instinct is to treat these as three separate fires, or to buy a “smart factory” system that promises to solve everything. Meridian does something else. It names the single decision that sits underneath all three: when do we stop the line to change a tool? If we do it too early, it wastes tooling and output. Made too late, it ships defects and fails the audit. And today it lives entirely in one retiring man’s ear.
By making that one decision observable, capturing the few signals that predict drift and the evidence that a change was justified, Meridian protects margin, satisfies the traceability mandate and begins converting tacit knowledge into a shared asset. One well-chosen decision answers all three pressures.
The blueprint that follows is what I call Frugal Architecture: building the leanest architecture that fully supports the firm’s decisions. Open where lock-in would be costly. Standard where interoperability matters. Simple enough that your own team can run and change it.
Frugality here is not “the cheap option”. Every system you adopt levies an ongoing tax in integration, updates, training, and attention, and that tax hits small teams with no slack hardest. Simplicity is a feature you must defend. Done well, a frugal architecture is cheaper to run, easier to trust, and, counterintuitively, more capable of change than the expensive alternative because there is less to unpick.
Here is the part that should give every SME leader confidence. The multinational has resources. You have velocity. Your decision lines are short and you are close to your customer. You can learn faster than a large competitor can schedule a steering committee.
But velocity is only an advantage when each cycle becomes retained learning rather than repeated fire-fighting. The SME that captures what it learns compounds. The one that re-solves the same problems merely survives. Size is not the game. Learning rate is.
Every Tuesday for the next 52 weeks, I will take one idea from The 99% Factory and make it usable on a Monday morning. The year follows the same four phases the book recommends for a real transformation:
Along the way, we will look closely at what Singapore’s national AI strategy means for advanced manufacturing, what multinational customers now ask of their suppliers, and how SMEs can use available support without taking on costs they cannot carry.
Before you read anything else, do one thing. Write down, in your customer’s own words, the single outcome that keeps or wins your most important account. Then write down the recurring decision that most affects it, and who makes that decision today.
That is your starting point. Not a platform. Not a model. A decision.