The Differentiation Playbook for Singapore Small Business

The Industry Map

Every industry in Singapore, its size, its trend, and where the data is good enough to act on

Picture a street in Singapore, and the mix of businesses on it. At one end, a gleaming glass tower that houses a global bank. In the middle, a shophouse where a lawyer and an architect share a floor. Along the pavement, a hawker centre, a few boutiques, a clinic, a tuition centre, a man who fixes watches. They all look like they are part of the same economy. They are not.

The bank tower, the shophouse professionals, and the hawker centre are in different industries, and the industries, not the businesses, are what this chapter maps. Because the single most important thing a small business owner can know about the terrain they are standing on is which industry they are in, what that industry is worth, how fast it is moving, and whether the public data is good enough to make a decision on — more than how hard they work.

This is the map of that terrain. It shows every industry in Singapore, what it is worth, its share of the economy, its trend, and whether the data on it is trustworthy enough to act on. It then asks the question the headline numbers never answer: which industries are structured so that a well-run small business can actually win, and which are structured to grind you down no matter how well you execute?

It does not recommend specific businesses. It maps the terrain, industry by industry, so that a business owner can see where they stand.

Chapter 1, Here's the map

How Singapore officially classifies its industries

Before any analysis of industries, you need the map of industries itself.

Singapore classifies every economic activity using the Singapore Standard Industrial Classification (SSIC 2020), the national standard, aligned to the international ISIC framework, maintained by Singapore's statistics authority and enforced by the company registrar at incorporation. Every single business in Singapore, at the moment it is born, must declare its principal activity under an SSIC code.

The structure is hierarchical, five levels deep:

• A section (a letter from A to U, twenty-one in total), say, C, Manufacturing.
• A division (two digits), say, 26, Computer, electronic & optical products.
• A group (three digits), say, 261, Electronic components & boards.
• A class (four digits), say, 2611, Electronic components.
• A sub-class (five digits), say, 26112, Wafer fabrication.

The twenty-one sections run the whole gamut of human economic activity: Agriculture & fishing (A), Mining & quarrying (B), Manufacturing (C), electricity, gas, and air-conditioning (D), water and waste (E), construction (F), wholesale & retail trade (G), transport (H), accommodation & food (I), information & communications (J), finance (K), real estate (L), professional services (M), admin & support (N), public administration (O), education (P), health (Q), arts & recreation (R), other services (S), households employing domestic workers (T), and extra-territorial organisations (U).

Two of these sections are astonishingly detailed: Manufacturing (C) breaks down into 355 five-digit sub-classes, and Wholesale & retail trade (G) into 168. That granularity is itself a signal, these are the two most closely tracked parts of the Singapore economy.

But why should a small business owner care about the bureaucratic code they declare at registration? Because the SSIC code is not a formality. It decides which statistical series your industry appears in, which grants you are eligible for, and how the government counts your sector. A business that picks a broad, or wrong, SSIC code disappears into an aggregate and becomes invisible, to the data, and to the policy that follows the data. Choosing the right code is the first act of positioning. It decides which map you appear on.

The scale, the size of the whole, and how it splits

The macro frame first, all of it from 2025, out of Singapore's statistics office. Two words appear on every page, so here is what they mean in plain dollars. GDP, gross domestic product, is the total value of everything produced in Singapore in a year, the whole economy measured in dollars. Value added is one industry's own slice of that, after the materials and services it bought from other industries are subtracted; it is the fairest measure of how big an industry truly is on its own.

• GDP at current market prices: S$789.5 billion, up 3.1 percent.
• GDP in chained (real) dollars: S$593.9 billion, up 5.0 percent, the real growth of the economy.
• Per capita GDP: S$129,194.
• Enterprise value added: S$714.7 billion.
• Total enterprises: 371,000.
• Total employment: 4.12 million (including foreign workers).

The services-producing industries generate over 70 percent of nominal GDP; the goods-producing industries, manufacturing, construction, utilities, generate the rest. This is the three-engine structure from the landscape chapter, now visible at the level of industries. There is the global engine, the industries that build and trade with the world, like manufacturing, finance, and wholesale trade. There is the government engine, public administration, education, health, the industries the state runs. And there is the domestic layer, the industries that serve people where they live, like retail, food & beverage, accommodation, and personal services. They are not just separate places. They are separate industries, with entirely different economics. Think of it as the machine room versus the shop front versus the state's own workshops: the engines make money for the whole country; the domestic layer takes that money from the people who hold it.

The industry-by-industry map

Here is the core table, the full sweep of every industry in Singapore, its value added in 2025, and its share of GDP. Each figure is verified directly from the statistics office's detailed national-accounts table.

A map that shows you everything is no help unless you know how to read it, so here are three habits to hold on to as you work through the table.

First, separate value from entry. The industries with the biggest numbers, wholesale trade, manufacturing, finance, are the engine's territory. Capital-intensive, meaning they need billions of dollars of machinery and scale before they turn a profit, dominated by multinationals and state-linked giants, and structurally closed to a typical small business. Do not read "biggest share of GDP" as "best opportunity." Read it as "most concentrated, hardest to enter."

Second, read the domestic tail, not the head. The industries a small business can actually enter, retail, food & accommodation, personal services, admin & support, health, education, arts, are each a small slice of GDP, but collectively they employ a huge share of the workforce. That is where the density is, where the competition is fierce, and where differentiation, not price, is the only durable way to survive.

Third, weigh the data against the opportunity. The industries with the best data, manufacturing, finance, wholesale, are the ones you cannot enter. The industries you can enter have good aggregate data but coarse competitive data. That inversion is the reason to keep this map: it tells you where the value is, where you can realistically play, and where you will have to do your own research, because the public data will not hand you the answer.

Here is the map.

Ownership of dwellings (S$29.7bn, a national-accounts adjustment for owner-occupied housing) is not an industry, it is excluded here. Figures are gross value added at basic prices; the shares of GDP sum to about 97%, the balance being taxes less subsidies on products.

Footnote on retail trade (S$9.0bn, 1.1% of GDP): this figure should be read with care, because it is sensitive to how wholesale and retail are split within SSIC section G. Wholesale trade (S$146.5bn) and retail trade (S$9.0bn) are both booked under section G, and the split between them depends on how each establishment classifies its principal activity. A large share of what a consumer thinks of as "retail", the big-format stores, the online sellers, the distribution that sits behind a shopfront, may be booked under wholesale or under another section. The S$9.0bn is the official retail-trade line, but it is a narrow measure of the retail activity a small shop actually competes in. [VERIFY: confirm the S$9.0bn is not an artefact of the wholesale/retail split within section G, and state the split either way.]

Footnote on health & social services (S$20.7bn) and education (S$18.4bn): these two lines sit partly inside and partly outside the government engine. The government engine's value added includes the public hospitals, the public schools, and the public administration that runs them, funded by the state, not by a market. But the same SSIC lines also include the private clinics, the private hospitals, the private tuition centres, and the private schools that compete in the market. The book leans on the government-engine distinction elsewhere, so it matters that these two lines are a blend: part state-funded, part market-funded. The private-tuition market alone is roughly S$1.8bn a year, a small slice of the S$18.4bn education line, which is dominated by the public school system. [VERIFY: state what falls inside versus outside the government engine for both lines.]

Now read it. The three giants, wholesale trade (18.6%), manufacturing (17.4%), and finance (13.2%), together account for almost half of Singapore's GDP. These are the engine-layer industries. Capital-intensive. Export-oriented. Dominated by multinationals. They are where the value is, but they are not where most small businesses live.

And the domestic-layer industries, retail, food, accommodation, the fragmented personal and admin services, are each tiny as a share of GDP. But together, they employ a huge share of the workforce. This is the density, the cap, the competition, of the landscape chapter, now quantified: the industries where most Singapore small businesses compete are individually tiny slices of the economy, in a market that is capped.

Here is the practical reading of the table: do not chase the big numbers. Wholesale, manufacturing, finance are the engine's territory, capital, scale, and corporate dominance make them hostile to a typical domestic SME (small and medium enterprise, the shops, clinics and centres with a handful of staff). The domestic-layer industries are small in GDP terms, but they are where the people are, where the premium demand of the engine's payroll flows, and where differentiation actually decides who survives.

Chapter 2, Which way is it moving?

A map of what is standing still is not enough. A business owner needs to know which way the terrain is tilting, where the demand is landing, and which industries are growing, stale, or declining. That is what this movement answers, in two connected reads: where the demographic waves land as industries, and which industries the growth numbers actually favour.

Where the demand lands, the bridge from the demographic chapter

The demographic chapter ended with a map of people: who is here, what they earn, what they spend, where their attention goes, and the four waves reshaping all of it. This chapter is the map of industries. A bridge between the two is the answer to the question, not a decoration every small business owner is actually asking: I have read about the ageing, the shrinking household, the collapse in births. Now show me where that money lands.

So here is the concrete answer, and it is more interesting than any headline number.

Take the oldest person on the island. The demographic chapter established that about 6.6 percent of Singaporeans over 65, roughly 52,000 people today, have a disability serious enough to need care, and that nearly all of them age in place, at home. Now follow that person to an industry. The care they need is a home-care aide who comes at twenty dollars an hour, a day-care centre that charges nine hundred to fourteen hundred a month, a meal delivery that puts food on the table, a monitoring service that tells a daughter her mother is safe — not an abstraction called "ageing." Each of those is an industry with a national-accounts line, a dollar figure for its size, an establishment count, a tally of how many operators are actually in it, and a growth rate. The demographic wave did not stay in the demographic chapter. It is sitting, right now, in the value-added of health and social services, which the Ministry of Trade and Industry (MTI, the government ministry that publishes Singapore's growth figures) describes as "resilient" even as other domestic industries slow.

Now take the childless couple. The demographic chapter showed that a young adult who does not have a child does not stop spending — they redirect the money that would have gone to a child into something else: the dog, the table, the travel, the experiences. Find that in the industry map and you find the food-and-beverage services line, the recreation line, the personal services line. The money did not disappear when the birth rate fell. It moved sideways, into industries, into lines of GDP that a business owner can name.

And take the shrinking household. A person who lives alone cannot share a dinner, a rent, a washing machine, a conversation. So they buy single servings, delivery, a laundromat, a pet, a streaming subscription. Every one of those is an industry, and the household wave is the growth line in food delivery, in convenience retail, in the personal services that keep a household of one running, not a sociological note.

That is the bridge. Here is the real shape of it:

Now here is the part that no single table shows, and the part that should stop a small business owner in their tracks. The demographic wave with the strongest tailwind, ageing, care, health, lands in the industries where the structure is also most favourable. Health and social services is the domestic-layer industry where regulation raises the barrier to entry, where trust stops pure price competition, and where the demographic tailwind is strongest. The people-map and the industry-map point to the same place, and that is the reason the series goes from "which wave" to "which industry" to "which slot."

The other way to read the bridge is the warning. The demographic demand does not land evenly. It lands hard in the industries that are hard to enter, health, care, premium professional, where regulation and trust gate the door, and it lands softly in the industries that are easy to enter and brutal to survive, retail, food, personal services, where the same demand produces a hundred competing shops. The bridge tells a small business not just where the money is but where the money is and the door is open for a positioned player. Those two do not always coincide, and knowing the difference is the point.

That is the value of the bridge, in one line: the people are becoming industries, and the industries that the people are becoming are the ones a positioned small business can enter, if they read which wave is their wave before they pick their door.

Growth, staleness, and decline, which industries are moving

Now the map that a business owner actually reads: not just which industry is big, but which is growing, which is stagnant, and which is shrinking. The MTI quarterly releases give the verified growth rates (2025 and 1Q2026, year-on-year):

Read the retail rows together and the trend is sharp. The retail aggregate is still growing, +5.0 percent in 2025, +4.0 percent in mid-2026, so the overall industry is not in decline. But the growth is not evenly spread. Computer and telecommunications sales are up nearly ten percent, the online share of retail is climbing toward a fifth, and the legacy formats are contracting, department stores down 5.7 percent, and F&B services, which boomed, are now softening (down 2.3 percent in June 2026). The retail industry is bifurcating between the formats the country is moving toward and the formats it is leaving behind.

The pattern that matters for a small business: the growth is at the two ends the map already flagged, the digitising engine (information, communications, online retail, computer/telecoms) and the ageing-adjacent services (health, care, premium professional). The staleness sits in the middle, the fragmented, price-war domestic services that grow slowly and cut hardest. The decline is concentrated in the legacy retail formats (department stores) and, for now, the F&B services that softened.

A business owner should read this as the demand-traffic map: growth industries are where the money is heading; stale industries are where it is already crowded and fighting over a slow-growing pie; declining industries are where the format itself is losing to a substitute (online, delivery, a newer way to serve the same need). Being in a growing industry does not guarantee you win, but it means the current is at your back. Being in a stale or declining one means the current is against you, and only a position that pulls you out of the commodity can save you.

Chapter 3, Where can a small business actually win?

The map told you what each industry is worth and which way it is moving. Now the question that matters most: inside these industries, where can a single owner-operator or small team actually win? This movement merges three things that the map used to treat separately, the specific slots, the industry detail, and the structural test of whether a well-run business can earn there. Read together, they answer one question: which industries are growing AND enterable AND structured so a positioned player can take share.

The structural test, where a well-run business can actually earn

Before the specific slots, ask the question the numbers alone cannot answer: is this industry set up so that a well-run business can actually earn a living, or is it set up to grind you down no matter how well you run it?

There is a way to test this that has nothing to do with business school. Think about a market you know well, say, the coffee shops in your neighbourhood, and ask five plain questions about it:

• How easy is it for someone else to start up and do the same thing? If anyone can register a business in a day and open up next to you, you can never relax. If it takes a licence, a lot of money, or a skill few people have, the crowd stays out.

• How much can the people you buy from squeeze you? If you depend on one supplier who can raise prices whenever they like, you have no room. If you have many suppliers, you are in charge.

• How much power do your customers have over you? If they can go anywhere and compare every price on their phone — they hold the power. If they need you specifically, you do.

• Could your customer get the same thing somewhere else, without you? If your service can be replaced by a cheaper or easier alternative, you are always one step from losing the sale.

• How hard do the shops around you fight? If every shop competes by cutting price, the whole street gets dragged down. If the shops are not fighting, each one content, doing the same thing for years, there is room for one that does it better.

Here is the twist that matters, and it is the whole point of this test. An industry where all five are hard, easy to enter, squeezed by suppliers and customers, easily replaced, fighting on price, is a brutal place to be if you are just like everyone else. But it is exactly the place where a business with a clear difference can win. The difficulty is the same difficulty every shop on the street faces. The one that stands out is the one that stops fighting on price and gives people a reason to come to it instead of the shop next door.

So do not read the industries below as a list of places to avoid. Read them as a list of places where the shops are all fighting on price, and where one operator with a clear difference takes the share of everyone around them. The difficulty is the sign that the incumbents are exposed, not a warning for you.

Here is the whole test across the domestic-layer industries at a glance, using the five plain questions. Read the last column first — it is the one that tells you where to look:

Two things to notice in that table. The industries where "Easy to start up," "customers hold the power," "can get it elsewhere," and "shops fighting on price" all say yes, retail, food, admin, personal services, are the brutal ones, and they are exactly the ones open to a business with a clear difference. The difficulty that grinds everyone down is the same difficulty that leaves every shop exposed. And the industries with "Hard to start up" or "Low" on customer power, health, professional services, are the friendlier ones, but they are harder to enter and the data is thinner. That is the whole structural picture in one table.

The industries, one by one

Here is where the map stops being a list of numbers and becomes a set of arenas. Each industry below is read the same way: what it is, how the structure cuts for a small player, and, for the domestic-layer ones a small business can actually enter, what the concrete opening is.

The engine's core, manufacturing, wholesale, finance, is where the value is, not where small businesses live. Manufacturing is the single most important goods-producing industry and the one with the best public data in the entire economy. Its value added in 2025 was S$132.8 billion, up 13.9 percent; its total output was S$479.1 billion, up 11.5 percent; it pays S$26.4 billion in remuneration; it is a fifth of the economy. The 2025 growth was led by electronics, transport engineering, and precision engineering, which more than offset declines in biomedical, general manufacturing, and chemicals. The data is the gold standard, value added, output, productivity all published by cluster, monthly and annually. But the strategic caveat is equally clear: manufacturing is dominated by multinationals and state-linked firms, is capital-intensive, and its value is concentrated in a small number of large firms. A domestic SME does not compete in manufacturing; it competes around it, as a supplier, a service provider, a niche player in a cluster the giants ignore. The data is excellent. The barriers to entry are brutal.

The same holds for wholesale trade (18.6% of GDP, the engine's trading arm, corporate territory, capital-intensive) and finance (13.2%, heavily regulated, dominated by the banks). These are where the value is; they are not where most small businesses compete.

Information & communications is the growth engine of the domestic layer, and for a small business, the most strategically important industry in the country. It is the fastest-growing domestic-layer industry, driven by "continued strong demand for IT and digital solutions." And it is where the differentiation opportunity lives: there is a roughly 48-point gap, a difference of almost fifty percentage points, between how fast small businesses adopt technology and how fast large businesses do, and it is being closed. The businesses that differentiate now own positions the late adopters cannot take. For a small operator, this is the arena where a single person with a position and a tool can out-execute a shop that has done the same thing for twenty years.

Professional services is the most interesting domestic-layer industry. Legal, accounting, management consultancy, architecture, engineering, research, advertising. The demand for specific, deep differentiation is outrunning supply by eight to one, management consulting registrations grew only 1.0 percent while business formations grew 8.5 percent. Competition is intense, and the state's free advisory centres anchor generic advice at zero. But the white space is in the deep, specific, niche position that the state does not provide, not in generic professional services. This is where a positioned small firm can genuinely earn, and where the incumbent advisors are the most vulnerable to a sharper, more specific competitor.

Health & social services is the friendliest domestic-layer industry for a small business that is willing to do its own homework. Regulation keeps new players out. Trust stops the fight over price. And the demographic tailwind, ageing, care, the silver economy, is the strongest in the economy. The catch is that the data is coarser, so the businesses that do their own research gain the edge.

Education is structurally mixed, with clear disruption pockets. A crowded field and a shrinking base, the fertility collapse. But the concentration of per-child spend creates premium pockets where a specialist takes the concentrated dollar from the generic 1,000+ centres.

Retail trade is hostile to businesses that compete only on price, which is why it is open to a business with a clear difference. Anyone can register a business in a day, so the field is wide open. Buyers can compare every price on their phone, so they hold the power. The same need can be met online or delivered, so customers always have another option. And the shops fight on price. The corporate reading is "escape." The small-business reading is the opposite: this is an industry of undifferentiated shops fighting on price, where one operator with a difference, a niche, a reason to be chosen beyond price, takes share from everyone around them. The field is open for you too. Use it.

Food & beverage is brutal, and brutally ripe for a business that stands out. High fixed costs (rent, labour), a capped market, intense competition, and a five-year survival rate under a quarter for the ones with no particular edge, meaning fewer than one in four new restaurants with no difference is still open five years later. But most of that competition is menus competing on price at the same ten dollars. The opening is the F&B business that is not competing on the plate, the one that owns a moment, a neighbourhood, a type of eater that people will cross the island for. The incumbents are not different from one another; they are interchangeable. That is the gap.

Admin & support services, employment agencies, security, cleaning, travel, is the "boring but real" SME economy. Most of it is a price commodity, but a commodity with complacent incumbents. A positioned operator (specialised security, white-collar recruitment, high-end cleaning) escapes the price war and takes the premium tier. The boring is exactly where a sharp operator with a clear word disrupts the unchanged incumbents.

Personal services, beauty, hair, repair, is the dense SME heartland. Retail is S$9.0 billion, food & beverage S$7.6 billion, accommodation S$5.6 billion, each tiny, together about three percent of GDP, yet employing a disproportionate share of the domestic workforce. Personal services looks hostile, but it is a field of undifferentiated shops with no protection. One specialist, one premium salon, one trusted repair brand, takes the premium tier. Because it is the least precisely measured industry, proprietary research is itself an edge.

Real estate is the industry that sets everyone's cost base. Not a typical arena for a domestic SME, but structurally powerful for the landlord and constraining for the tenant. A small business cannot change the real estate structure; it can only choose which part of it to pay, downtown premium or heartland mass. That choice, made consciously, is a positioning decision.

Transportation & storage, the logistics spine. Land, water, air, warehousing. A domestic-layer industry with real SME participation, though capital-intensive at the top.

The slots, where the obvious wins are

The structural test told you which industries are worth entering. The slots are the specific openings a small business can walk into, the word, the service, the niche no incumbent owns. When this chapter says SMB (small and medium business, the shops, clinics and centres with a handful of staff) or OPC (one-person company, a business run by a single owner-operator), both are exactly the reader this whole series is written for, the small operator who does not have a chain's capital or a corporate team. Read it as three columns. Wave is the demand (from the demographic chapter). Industry is where it shows up. Slot is the specific opening a small business can take. A few of these slots are so clearly exposed that they are the obvious wins, and here they are:

Each of these is a slot, not a business. A slot is where a position can live. And the pattern across all of them is the same: the winning slots sit in industries the big players find too small or too messy, serving a demand the demographic waves are making bigger every year, with an incumbent who has not changed in a generation.

Why these are the obvious wins for a small business, an SMB or a one-person company, specifically:

• They are capital-light. No manufacturing line, no banking licence, no office tower. The cost is a person, a position, a name, a way of doing one thing well.
• They are trust-driven. The buyer wants a named, vouchable operator, not a platform. That is the moat from the first chapter, and it favours the small player who can be a person, not the incumbent who is a brand.
• The incumbents are complacent. A fragmented industry of small shops that all do the same thing is exactly the arena where one sharp operator with a clear position eats the share of everyone around them. The barriers are low, but they are low for you too.

Do not read this as "the industry is brutal, stay away." Read it as the opposite: the industries that look brutal from the outside, through a corporate lens, are the ones where a positioned small business can take share from shops that are all doing the same thing. The very things that make those industries hard, easy for anyone to enter, customers who compare prices, shops that fight on price, are the opportunity, because it means every incumbent is as exposed as you are. The difference is you have a clear position and they do not. The slot is where your position makes the crowd irrelevant.

The slots above are the obvious ones. The demographic chapter and this map give you the method to find the rest: take a wave, find the industry it is growing, and look for the slot the incumbents are too slow or too generic to take.

Survival and entry cost, the numbers the map does not usually show

The structural test tells you which industries are worth entering. It does not tell you how much it costs to enter, or how likely you are to survive once you are in. Those two numbers matter as much as the structure, and they are thinner in the public data. Here is what is known, and where the gaps are.

The national survival figure is a single number, and it hides the industry split. About 49 per cent of companies registered in 2020 were still active six years later. That is the whole-economy average. The split behind it is the more useful number: the high-barrier sectors, those needing more capital or serving business buyers, show survival rates of 70 per cent and above, while the low-barrier consumer trades, retail, food, gig-style services, see most entrants fail. The gap between the two is the gap between an industry that filters entrants by capital and one that filters them by survival of the fittest.

Sector-level survival data is the gap, and the closest public proxy is sector-level cessation counts. Singapore publishes annual business-cessation counts by industry, through SingStat's table builder (the "Cessation of All Business Entities by Industry" series). These show which sectors close the most businesses each year: in 2025, retail trade saw about 6,800 business cessations, wholesale trade about 10,000, professional services about 9,600, and food and beverage services about 3,100. That is useful — it confirms which arenas are dense with closures. What it is not is a clean five-year survival cohort by industry: a number that follows one year's entrants forward and reports what share survived. That figure does not exist in the published data. The 49 per cent national survival rate, the high-barrier/low-barrier split, and the sector-level cessation counts are the best the public data supports. The granular version — the five-year survival rate for a hawker stall versus a tuition centre versus a home-care provider — does not exist in the published data, and this book will not invent it.

Entry cost is knowable as a range, and the range is wide. The picture of what it costs to open a small business in Singapore, by industry, with wide ranges and confidence labels:

These are wide ranges with confidence labels, not precise figures, because Singapore does not publish a standard entry-cost table by industry. They are more useful than silence: they tell you that the capital-light arenas, professional services, home care, tuition, are the ones where a person with a skill and a position can start, and that the capital-heavy arenas, retail, F&B, salon, are the ones where the rent and fit-out are the first filter.

Owner earnings are the least-published number of all. The map shows value per worker by industry, but that is not the same as what an owner takes home. A hawker's value per worker is low, but the owner's take is what is left after the stall's costs, and that is not published. The statement is that owner earnings by industry are not in the public data, and the value-per-worker figures in this book are the closest proxy, with the caveat that they measure the business's output per worker, not the owner's income. [VERIFY: check whether any source publishes owner-operator earnings by industry; if not, mark as a gap.]

The point of these numbers is to make the entry decision honest, not to scare an entrant. The industries with the lowest entry cost and the thinnest survival data, the fragmented consumer trades, are exactly the ones where a position matters most, because the crowd is the thing that kills you, and a position is the one thing that lifts you out of the crowd. The industries with the highest entry cost and the best survival, the capital-heavy, licence-gated ones, filter entrants by money, and a position matters less because the barrier does the filtering. The map's advice holds: if you are entering a low-barrier industry, you are betting on your position to survive the crowd. Enter it knowing that is the bet you are making.

Chapter 4, What can you trust, and what does it mean?

The data, which industries can you act on?

This section separates a useful map from a decorative one. Not all Singapore industry data is equal. Here is the assessment of which industries have public data good enough to act on, and which are data-poor.

Excellent data, you can act on it. Manufacturing (value added, output, productivity by cluster, monthly). Finance (extensive statistics). Wholesale (detailed trade figures). Information & communications (InfoComm statistics, adoption data). Professional services (services + registration data). Transportation. Real estate (property and rental indices).

Directional data, not precise. Health & social services (aggregate published, but segment-level is under-quantified). Admin & support (aggregate; sub-segments not separated). Education (aggregate; the private-tuition market not precisely measured). Arts & recreation (aggregate; thin). Other personal services, beauty, hair, repair, laundry, the least precisely measured industry in the entire economy. Utilities (aggregate, state-dominated). Agriculture (tiny).

Now read those two lists together, because they reveal the whole point.

The industries with the best data, manufacturing, finance, wholesale, are the ones a small business cannot enter. The industries where a small business actually competes, retail, food, accommodation, admin, personal services, health, education, have good aggregate data but coarse competitive data.

This is not a reason to stay away from the domestic-layer industries. Far from it: the industries with the strongest tailwinds, ageing, care, health, are precisely the ones where the public data is thinnest. So the businesses that do their own primary research gain an information advantage the data-poor competitors do not have. In a data-poor industry, proprietary research is itself a differentiation.

The gaps, closed with the data that is available

The national-accounts map stops where the aggregation stops. But the gaps it leaves are not empty. Below are the under-measured industries, sized from the online data that does exist, market research, industry registries, platform counts, and parliamentary answers. None of these is a national-accounts figure. Each is directional, assembled from available sources, and flagged as such. But for a business deciding where to play, a directional number from the real world beats a precise blank.

Personal services, beauty, hair, repair, laundry, is not actually un-measurable. The most often-repeated gap in this map has a number hiding in plain sight. The Singapore industry landscape (SkillsFuture / SSG) counts about 5,540 establishments in the Personal Care, Beauty & Hair Dressing Services sector, with operating receipts of around S$1.37 billion a year. That is the single most useful figure in the whole domestic layer — it tells a business the real size of the arena it is competing in. It is not in the national accounts as a clean line, but it is available, and it is honest about its source: a government skills agency's industry profile, not a precise census. Caveat: it aggregates beauty, hair, and dressing, so a nail-only or hair-only entrant must estimate their slice. But the total is real, and it is over a billion dollars.

The tuition industry is better measured than the national accounts suggest. The official numbers understate it because MOE does not track the private industry's total value, but the industry itself is visible from several angles. Households spend about S$1.8 billion a year on private tuition (HES 2023). Around 1,000+ tuition and enrichment centres are registered with MOE, a number that has grown steadily and is now higher than the count of all primary and secondary schools combined. Seven in ten parents send a child for tuition, and eight in ten primary-school children pay for it. The industry has more registered centres than the national education system has schools, which is itself a sentence every tuition operator should read twice. Caveat: the centre count misses the tens of thousands of freelance tutors who are not registered, so the true supply is larger; the household-spend figure misses corporate and online providers. But the scale is no longer "unknown" — it is a billion-dollar-plus, thousand-plus-centre, majority-of-families industry.

Eldercare / home-care providers are countable from the service registries. The state is building this market deliberately, and it publishes the counts. The plan is to roughly double the number of eldercare centres to 220 by 2025 (MOH), and to double nursing-home beds. Active Ageing Centres have grown from 119 to 223. The licensed home-care and day-care providers are listed in the AIC's care-services directories. The count is not a single published "market size," but it is a countable supply side: the number of centres and beds is available, and it is rising on state funding. That is the caveat-and-use figure for a home-care operator: the supply is growing, and the state is the reason.

The retail competitive field is readable from the platform/aggregator data. The national accounts give retail as an aggregate; the fine-grained competition is visible elsewhere. Establishment counts (6,355 restaurants, 837 fast-food outlets, 5,540 beauty businesses) give density. The e-commerce and delivery platforms give share, Shopee and Lazada dominate, cross-border takes over half of online spend, online is ~a fifth of retail. A retail entrant can reconstruct its competitive field from these fragments even though no single table publishes it.

Ownership of dwellings, the S$29.7 billion imputation for owner-occupied housing, is not an industry. It is a statistical adjustment, and the map deliberately excludes it.

The silver-economy market size, the one genuinely open number, is now closed by construction. The US$72.4 billion headline is senior total spending power, not a market. The demographic chapter built the addressable paid care-services layer, home care, day care, residential care, meals/nutrition, safety/monitoring, at roughly S$0.28bn–0.85bn a year, anchored by a verified care-need base of 6.6% of 65+ with moderate-to-severe disability (NUS/MOH RHS) and Duke-NUS's finding that ~75% of senior care is informal/unpaid family care. It excludes the migrant-domestic-worker channel and the wellness/active-ageing spend of the healthy majority, which would add a multiple. Any business that extends it, or builds the adjacent wellness layer, holds an information advantage no headline provides.

The gaps are now filled with estimates. None of these numbers is a national-accounts figure; each is assembled from available public data, a skills-agency industry profile, an MOE registration fact, a parliamentary household-spend figure, a service-registry count. They are directional and caveated. But they are no longer blank. And in a map where the official data stops early, a directional number built from the available sources is worth more than a precise empty space.

What this map means for a business

The map is the answer to "where can I even play?" But it answers it in the opposite way to how a corporate strategist would. A corporate reader sees an industry where anyone can enter, customers hold the power, and shops fight on price, and concludes "avoid." A small business owner should read the same industry and conclude "enter", because those are exactly the conditions under which every shop is exposed, and the tool that would once have been a large company's advantage is now available to one person with a clear position and a machine.

Put the whole structural test across the domestic-layer industries, and the picture is clear. Health & social services and specific professional services are the friendliest, regulation keeps new players out, trust stops the fight over price, and the demographic tailwind is behind them. Retail, food, admin, and personal services are hostile to anyone who competes on price, and therefore open to anyone who does not. The very difficulty that grinds down a new entrant is the same difficulty that leaves every shop around you exposed: easy to enter, customers who compare, shops fighting on price. A business with a clear difference takes the premium tier, the niche, the slot the interchangeable shops have left empty. Education has pockets of premium in a shrinking base. Real estate is powerful for the landlord, constraining for the tenant.

Your position is the open slot in the industry you can actually enter, the word nobody in that industry owns yet, the niche the incumbents are too slow or too generic to take. The map does not hand you the word. It tells you which industries are worth looking for one in, and it tells you where the incumbents are weak enough that your word will actually take share.

There is one more force to read into this map, and it is the newest in the economy. AI does not flatten these industries — it gives the small player the tool the incumbents do not have. The industries that look "boring" or "brutal", retail, food, admin, personal services, the fragmented professional tiers, are precisely the ones where a single operator with AI can out-execute a shop that has done the same thing for twenty years. AI does the execution work, the scheduling, the quoting, the follow-up, the bookkeeping, the first draft, the analysis, at a cost and speed the incumbent's headcount cannot match. The trust and the judgement and the position remain with the small operator; AI just removes the reason the incumbent could compete on scale. The map is a list of industries where a positioned, AI-armed operator disrupts the incumbents, not a list of industries to avoid. That is the subject of the AI chapter, two chapters ahead; the next chapter, the brand map, names the incumbents it is worth disrupting.

That is why the series moves from "which wave" to "which industry" to "which slot." The demographic wave tells you which demand is growing. The industry tells you where that demand lands. The slot is the opening you can walk into. The next chapter, the brand map, names who already owns each industry, and shows the word no one has claimed. Then the AI map tells you which of those slots a small operator can take faster, cheaper, and better than the incumbents, because they have the machine and the incumbents do not.

Working the map, three real-shaped industries, read honestly

A map is only useful if you can read it on a real business. It is worth walking the structural test through three real-shaped industries, so you can see how the map turns into a decision. Each is a different answer to the same question: is this a ground a small business can win on, and how?

The home-care operator. The demographic chapter showed the care wave is real and growing: the plan is to roughly double eldercare centres, and the senior share of the population is rising every year. The industry map adds the competition: regulation keeps new players out of the licensed layer, trust stops the fight over price, and the state is deliberately building the supply. That is a ground that is friendlier to a small operator than almost any other in the domestic layer, the wave is behind it, the competition is gated, and the customer buys on trust and referral. The small business that can own a slot in that ground, a specialist care team, a neighbourhood champion, a trusted referral name, is entering one of the few industries where the map says the ground is genuinely good.

The hawker and the small F&B operator. F&B is the hardest ground in the map. The value of the average F&B business is the lowest of any sector, the survival rate is brutal, fewer than a quarter of F&B businesses make it five years, and the competition is dense and undifferentiated. The structural test says this industry is hostile to anyone who competes on price, because everyone is competing on price. And yet the map does not say "avoid F&B"; it says "avoid being another interchangeable F&B shop." The hawker who owns a dish, a neighbourhood, a word the customer can repeat and refer is doing the one thing that breaks the hostile pattern: not competing on price, but owning a position. The difficulty of the industry is exactly what leaves the interchangeable shops exposed and the positioned one strong.

The small professional services firm. Professional services sits on the friendly end of the map, the value per worker is high, the customer buys on trust, and the demographic wave of foreign professionals brings new demand. The map says this is a ground where a small firm can win, if it can answer one question: what is the slot? The generalist accountant or consultant is interchangeable and fights on price. The specialist, the firm that owns one industry, one type of client, one problem, is the firm the customer recalls and refers. The industry is friendly; the position is the work.

Read these three together and the pattern of the map is clear. The map does not tell you the industry to be in; it tells you how to read the industry you are in. The home-care operator is in a friendly ground and still must position. The hawker is in a hostile ground and can still win by positioning. The professional is in a friendly ground and still needs the slot. The map is the ground; the position is the move; and every ground, friendly or hostile, rewards the positioned operator and punishes the interchangeable one.

Sources & confidence

All figures are confidence-labeled in the text. Confidence labels: High = primary source (SingStat, MTI, MOM, DOS) or multiple independent sources; Moderate = secondary or single-source; Low = derived estimate. The industry value-added and share-of-GDP figures are verified directly from SingStat's detailed national-accounts table (M015731, 2025). Growth figures are from MTI's 2025 and 1Q2026 releases (overall GDP +5.0% in 2025; services producing +4.3%; info-comms/finance/professional +4.2%; accommodation/real estate/admin/other services +3.1%; manufacturing +8.8%; wholesale-retail/transport collective +6.7%). Health & social services and education "remained resilient" (MTI 1Q2026). Retail growth +5.0% (2025) and +4.0% (Jun 2026); F&B services +3.1% (2025) but −2.3% (Jun 2026); computer & telecommunications +9.8%; department stores −5.7% (SingStat Monthly Retail Sales & F&B Services Index). Manufacturing cluster detail from MTI 1Q2026. Establishment counts from SingStat. Competitive registration data from ACRA.

Directional industry estimates (assembled from available online data, caveated): Personal Care, Beauty & Hair Dressing Services, ~5,540 establishments, ~S$1.37bn operating receipts (SkillsFuture/SSG industry profile, Moderate); private tuition, ~S$1.8bn/yr household spend (HES 2023), 1,000+ MOE-registered centres, 7-in-10 parents send a child for tuition (MOE parliamentary answer; SmileTutor/TODAY secondary, Moderate); eldercare/home-care, ~220 eldercare centres target by 2025, 223 Active Ageing Centres, doubling nursing-home beds (MOH/AIC, Moderate); e-commerce platform share and cross-border share (secondary).

Where a figure is directional, it is flagged in the text. All figures reflect data as of August 2026 and should be refreshed at publication.