Opening a childcare centre in Singapore, decoded.
A licensed trade with one of the stickiest customers in the country — a child can stay from 18 months to K2, nearly five years. You legally cannot open one without an ECDA licence, and the premises are government-gated. That is a real moat. The catch is the other side of the same coin: the affordable tier is price-capped by the state, teachers take roughly two-thirds of every dollar, and Singapore is now building places faster than it makes babies. Here is the picture the “how to start” blogs skip — the real cost, the licence, the fee caps, the occupancy maths, and where a new centre actually wins.
S$610 / 650
AOP / POP monthly fee cap (ex-GST, 1 Jan 2026)
~67%
of revenue is manpower (the listed proxy)
0.87
resident TFR in 2025 — a record low
<90%
utilisation in every planning area
The MOAT Score: is a childcare centre worth building?
Before the details, our one-number read. The MOAT Score grades a sector's economic quality on the value-investing lens of Graham, Buffett and Munger — four pillars (Margin, Operating moat, Appetite, Treadmill), each out of 25. At 60/100 (C), childcare lands in the middle — below tuition and dentistry, above cafés and boutique fitness. The reason is a rare split personality: a genuine licence moat and deep switching costs, held in check by a legal fee cap on most of the market and a shrinking pipeline of children.
Needs a real edge
Only with a real edge.
How the score is built
The MOAT Score sums four pillars — each scored 0–25 — from the value-investing lens of Graham, Buffett and Munger. No black box: here is the working.
Margin
12/25Does the average operator actually keep money — real net margin and return on the capital tied up?
This sector: real but thin — ~3.8%SourceMindChamps PreSchool FY2025 results, SGX, Feb 2026 net at the only listed proxy (and disposal-inflated; the underlying operating business is ~breakeven), with manpower around two-thirds of revenue and the affordable tier fee-capped.
Buffett 1979 — “a high earnings rate on equity capital… without undue leverage”; 1986 owner earnings.
Operating moat
16/25Pricing power and a durable competitive advantage — can a typical operator raise prices and have customers shrug?
This sector: the standout: a statutory ECDA licence you cannot operate without, government-gated premises, and deep enrolment switching costs (a settled child can stay up to ~5 years). The debit is that pricing is capped by law for the ~65%+Sourcegov.sg — share in govt-supported preschools, 2024 in AOP/POP/MOE places — a franchise muzzled, in Buffett's sense.
Buffett, FCIC 2010 — pricing power is “the single most important decision”; 1991 franchise; 2007 moat.
Appetite
19/25Demand durability — steady, recession-resistant repeat demand vs fragile, discretionary or faddish.
This sector: near-structural, non-discretionary demand: ~92%SourceMSF — preschool participation, 3–6 year-olds, 2021 of young children are enrolled and dual-income is the norm. The ceiling is demographic — a record-low 0.87SourceSingStat — resident TFR 2025, May 2026 fertility rate caps the cohort, even as per-child demand stays firm.
Graham, Security Analysis Ch.2 — inherent stability “derives from the character of the business”.
Treadmillinverted · less is better
13/25Capital intensity and structural drag — rent, churn, fashion, discounting. Scored inverted: less treadmill, more points.
This sector: heavy leasehold fit-out and a ratcheting manpower treadmill (ECDA keeps raising salary floors against a real educator shortage) — but the enrolled base is sticky, not re-acquired daily, so it avoids the brutal churn treadmill of a café or studio.
Buffett 2007 — “the worst sort of business… requires significant capital… Think airlines.”
M + O + A + T, out of 100
The MOAT Score is a transparent SGAI judgement on a sector’s economic quality through a value-investing lens — not a verdict on any individual business, and not a comment on an owner-operated livelihood (a sector can score low on capital returns yet work as a job).
The read: a real moat, with the state holding the price — and the demographics holding the ceiling
Start with what makes childcare different from almost every other SME a founder researches. A café, a studio, a marketing agency — anyone with capital can open one tomorrow. A childcare centre, you cannot. You need an ECDA licence under the Early Childhood Development Centres Act, your premises must pass a licensing visit, and most centre space is released through government tender (HDB void-decks, state sites). That is a moat in the precise Buffett sense — a structural barrier that stops the next entrant from competing your returns away.
And the customer is gloriously sticky. A child enrolled at eighteen months can stay through nursery, K1 and K2 — close to five years in one centre. Parents form their shortlist years before Primary 1, pay a registration deposit, and almost never move a child who has settled, made friends and bonded with a teacher. Siblings follow siblings. The recall is automatic: re-enrolment each January is the default, not a decision. That is genuine switching-cost lock-in — the kind of demand a value investor pays up for.
So why isn't this a wonderful business? Because the state sits on both ends of it. On price: roughly two-thirds of all preschoolers are in government-supported places, and those places are fee-capped by law — from 1 January 2026, S$610/month for full-day childcare at an Anchor Operator, S$650 at a Partner Operator (before GST), and those caps have been cut repeatedly. Buffett's definition of a real franchise requires one thing childcare's mass market lacks: that it is “not subject to price regulation.” The licence is a moat; the fee cap is a muzzle on the same animal.
On cost: the one input you cannot do without — qualified educators — is structurally scarce, and ECDA keeps raising the salary floor to keep them in the profession. Manpower runs about two-thirds of revenue. A capped fee on top of a rising wage floor is a vice that closes a little each year — which is exactly why the affordable tier needs recurrent government funding to be viable at all.
Then there is the ceiling. Singapore deliberately built enough full-day places for every resident child aged three and up — while the birth rate fell off a cliff. The resident total fertility rate hit a record-low 0.87 in 2025, and resident births fell about 11% in a single year. Median utilisation is already 69% for infant care and 74% for childcare, and no planning area is above 90% full. The era of “just get a licence and fill it” is closing. The founder's real question has shifted from can I open one? to can I win occupancy in a softening, capped, staffing-starved market?
Two markets, one capped by the state
Childcare isn't one pricing business — it's two. The affordable tier is fee-capped by ECDA (price-taker, by law); only the uncapped premium tier is real price-maker work. The mix decides everything.
- MOE Kindergarten (half-day)S$150
state-run, ~S$150 cap — the affordability anchor
- Anchor Operator full-dayS$610
capped by law (ex-GST, 1 Jan 2026) — price-taker
- Partner Operator full-dayS$650
capped by law (ex-GST, 1 Jan 2026) — price-taker
- Private full-day (typical)S$1650
uncapped
- Premium / internationalanchorS$2500
uncapped — the only real price-maker tier
Source: ECDA AOP/POP fee caps (eff. 1 Jan 2026) + MOE Kindergarten fees + DollarsAndSense private-fee survey (Feb 2026; indicative, ex-GST)
The map: ~1,900 centres, ~220,000 places — and the “market size” dollar number is junk
The honest, primary numbers: ECDA's live centre listing held about 1,900SourceECDA “Listing of Centres”, data.gov.sg, Jun 2026 licensed childcare centres and kindergartens, and full-day preschool places have grown to 220,000+SourceMSF Committee of Supply 2026, 2026 — now enough for every resident child aged three and up. Ignore the dollar “market size” reports entirely: the only figures that exist are global / APACShaky figure — treat with cautionMordor, Grand View, IMARC et al.These aggregators publish a GLOBAL (~US$260–350bn) or APAC childcare figure and disaggregate it top-down. None is a measured Singapore number — paywalled, methodology-opaque, and they conflate day-care, babysitting and edtech. Do not present any as the size of the SG market. models dressed up as Singapore. There is no credible Singapore-specific dollar market size; the defensible anchors are the place count (~220k), the centre count (~1,900) and government spend (~S$1.8B/yr as of 2021, since stated to be rising). The structurally important fact: government-funded operators dominate the affordable market — the five Anchor Operators alone run hundreds of centres — so a new private centre competes at the edges, not the centre.
~1,900
ECDA-licensed centres
data.gov.sg, live (Jun 2026)
220,000+
full-day preschool places
MSF COS 2026 — enough for every child 3+
~65%+
in govt-supported preschools
AOP / POP / MOE — target 80%
69% / 74%
median utilisation (infant / childcare)
MSF, end-2024 — system-wide headroom
The ceiling: a falling birth rate
Unlike a demographic tailwind, childcare faces a demographic headwind. Singapore built places faster than it makes babies — the resident fertility rate keeps setting record lows, capping the cohort the whole sector competes for.
2023—a record low — and still falling
2025—0.87 — a new record low (SingStat, May 2026)
Source: SingStat resident total fertility rate (2015, 2023–2025 reported; 2020 indicative of the trend), released 6 May 2026.
The players: who you're really studying
The market splits three ways: the government-funded mass-market giants (the Anchor Operators — capped fees, big scale, recurrent funding), the funded-but-competitive Partner Operators, and the uncapped private / premium tier where the real pricing power lives. The one with public numbers — MindChamps — is your single best window into the cost structure, even if it is a thin proxy for SG revenue.
The three ways to compete in Singapore childcare
| Player | Model | Licence moat? | Pricing power? | Signal |
|---|---|---|---|---|
| PCF SparkletotsAnchor Operator (largest) | Capped mass-market at scale (360+ centres) | The fee-capped, funded incumbent — ~40,000 children | ||
| NTUC My First SkoolAnchor Operator | Capped mass-market (~165 centres) | Funded scale; also runs premium Little Skool-House | ||
| Star LearnersLargest Partner Operator | Capped, funded, 5-yr POP appointment (~44) | POP is the funded route — but only 380 slots, by tender | ||
| MindChamps PreSchoolSGX-listed (premium) | Uncapped premium (~40 SG centres) | ~S$2,500/mo fees; the only public P&L (see below) | ||
| EtonHouse / Busy BeesPremium / international groups | Uncapped premium + an AOP arm (E-Bridge) | Where the price-maker margin lives |
PCF Sparkletots
Anchor Operator (largest)
- Model
- Capped mass-market at scale (360+ centres)
- Licence moat?
- Pricing power?
- Signal
- The fee-capped, funded incumbent — ~40,000 children
NTUC My First Skool
Anchor Operator
- Model
- Capped mass-market (~165 centres)
- Licence moat?
- Pricing power?
- Signal
- Funded scale; also runs premium Little Skool-House
Star Learners
Largest Partner Operator
- Model
- Capped, funded, 5-yr POP appointment (~44)
- Licence moat?
- Pricing power?
- Signal
- POP is the funded route — but only 380 slots, by tender
MindChamps PreSchool
SGX-listed (premium)
- Model
- Uncapped premium (~40 SG centres)
- Licence moat?
- Pricing power?
- Signal
- ~S$2,500/mo fees; the only public P&L (see below)
EtonHouse / Busy Bees
Premium / international groups
- Model
- Uncapped premium + an AOP arm (E-Bridge)
- Licence moat?
- Pricing power?
- Signal
- Where the price-maker margin lives
The only public window: MindChamps PreSchool
MindChamps (SGX) is the only listed pure-play, so it is the cleanest read on what a centre actually earns. FY2025: S$60.4M group revenue, net profit ~S$2.3M (~3.8%), with employee compensation around 67% of revenue and rent ~12%. Read it with two caveats. First, that thin net profit leaned on ~S$7M of one-off disposal gains — the underlying operating business is closer to breakeven. Second, Singapore is mostly a franchise/royalty book for MindChamps (Australia is ~68% of revenue), so it is a weak proxy for SG centre revenue but a good one for the cost ratios — and those ratios say manpower is the business.
The customer: dual-income parents, choosing on proximity then price-after-subsidy
The buyer is the dual-income household — among prime-age women (25–64), labour-force participation has reached ~80.5%SourceMOM Labour Force in Singapore 2025, 2025, which is why full-day, not half-day, care dominates demand. Parents shortlist on proximity first (home, work, or grandparents), then fee-after-subsidy, then curriculum and “school readiness.” Subsidies attach to the family, not the centre — so a private-centre family still collects the Basic (and means-tested Additional) subsidy; they just face an uncapped fee.
What a working family actually pays, by tier
Indicative monthly out-of-pocket for a full-day childcare place after a working-mother's Basic Subsidy (S$300/mo). The subsidised tiers are nearly free; the premium tier is where the real money — and the real margin — is. Income-eligible families get more on top.
- Anchor Operator (S$610 cap)15%after Basic Subsidy; lower-income families pay close to S$0 — the affordability play
- Partner Operator (S$650 cap)17%after Basic Subsidy — capped, funded, competitive tender
- Private / premium (~S$1,650+)67%after Basic Subsidy on an uncapped fee — the price-maker tier
Source: SGAI illustration using ECDA fee caps + the S$300 working-mother Basic Subsidy (ECDA subsidy table, 18 Nov 2025), before GST and any means-tested Additional Subsidy. Indicative of the out-of-pocket structure, not a measured per-family figure.
The kiasu lever: “school readiness”
Parents say they want holistic, play-based development; many buy Primary-1 readiness — phonics, literacy and numeracy, and bilingual / Chinese enrichment. That anxiety is the single biggest reason a family pays double for a private centre, and it is the cleanest way for a new operator to escape the fee cap. The state pushes holistic development; the wallet still rewards a credible readiness story.
The moat you can't buy: stickiness
Once a child is settled, the family almost never moves them — the disruption to a young child, the lost deposit, the new waitlist and the sibling already enrolled all bind them in. Up to five years of fees per enrolment, re-booked automatically each January. No SG retention rate is published, so we assert this qualitatively — but it is the heart of the moat, and the reason occupancy, once won, tends to hold.
The economics: the licence is cheap, the teachers are everything
Here is the inversion founders miss. The ECDA licence costs almost nothing next to the build. What costs real money is the fit-out and the people. There is no clean Singapore startup figure — the “~S$800,000+” you'll see online is a consultancy estimate, not an official number — but it is dominated by renovation to meet ECDA floor-area, safety and outdoor-play rules, the deposit, and several months of payroll before the places fill. Then the biggest operating line, by far, is the teachers: roughly two-thirds of revenue, against ECDA salary floors that keep rising. Against a fee cap, that is the whole tension — and it is why occupancy, not the fee, decides whether you make money.
All-in cost to open a centre
Renovation/fit-out to ECDA standards (the bulk), deposit, equipment and several months of pre-opening payroll. The “~S$800k” figure circulating online is a consultancy estimate with no itemised basis — Singapore publishes no clean per-centre startup cost, so treat any precise number as folk knowledge and verify every line.
Source: Consultancy/advisory estimates (e.g. Piloto Asia, 2025) — flagged as a folk figure, not an official statistic
Median childcare utilisation is ~74% (infant care ~69%), and no planning area is above 90% full. Occupancy is the swing factor in the whole P&L — and in a falling-birth market, it is getting harder, not easier, to fill.
MSF — Update on Infant Care & Childcare Places, end-2024
Where the childcare dollar goes
A typical SG centre, modelled per S$100 of fees. Manpower — not rent — is the dominant line, anchored on MindChamps' ~67% (the listed proxy). What's left is a thin margin that only appears at high occupancy.
- Teachers + support staff + CPF−65%35% left
the business is its educators; ECDA keeps raising the salary floor · MindChamps FY2025 (~67% of rev) — SGX
- Rent + occupancy−12%23% left
low for HDB void-deck, far higher for private/mall sites · MindChamps FY2025 (~12%) — SGX
- Food + consumables−6%17% left
meals, learning materials, nappies (infant care) · SGAI estimate — no clean SG figure
- Depreciation (fit-out)−8%9% left
the heavy renovation, written down over the lease · MindChamps D&A proxy — SGX
- Other fixed (utilities, insurance, software, marketing, licence)−6%3% left
What the owner actually keeps
Verdict: A thin 3% — survivable, not comfortable; one bad assumption flips it negative.
Illustrative model: MindChamps' audited cost ratios (manpower ~67%, rent ~12%) anchor the lines; the rest are SGAI estimates where no clean SG figure exists. Net margin is thin (the listed proxy ran ~3.8%, disposal-inflated) and only materialises at high occupancy. Capped AOP/POP centres rely on government funding to clear breakeven. Not financial advice.
Childcare & preschool
Would a value investor own the average operator here?
A value investor would want a well-run premium centre, not the average one. The moat is real — a statutory licence plus near-five-year enrolment stickiness — but the affordable market is fee-capped by law and the cohort is shrinking, so the average operator is squeezed between a capped fee and a rising wage floor.
Two-thirds of the market is fee-capped by ECDA (AOP S$610 / POP S$650, 2026); only uncapped private/premium centres are real price-makers. Buffett: a franchise must be “not subject to price regulation.”
Buffett 1991 — a franchise “not subject to price regulation”; FCIC 2010 — pricing power is “the single most important decision”
A statutory ECDA licence you cannot operate without, government-gated premises, and deep enrolment switching costs (a settled child stays up to ~5 years). Strong — but muzzled by the fee cap on most of the market.
Buffett 2007 — an enduring moat protects returns on capital
Solid at high occupancy in the premium tier; thin-to-breakeven in the capped tier without government funding. Heavy fit-out means ROIC builds slowly.
Buffett 1979 — a high earnings rate on capital, unleveraged
Leasehold fit-out to ECDA standards is the big sink; to add capacity you build and staff another centre.
Buffett 2007 — the worst business needs much capital, earns little
Non-discretionary, subsidy-railed, ~92% participation — but a record-low birth rate caps the cohort the whole sector chases.
Graham, Security Analysis Ch.2 — inherent stability is qualitative
Either a funded AOP/POP appointment run at scale and high occupancy, or an uncapped premium/niche centre (bilingual, school-readiness, infant care, inclusion) that escapes the fee cap — with occupancy and educator retention as the two metrics that actually decide profit.
Assessment uses the value-investing lens on SG childcare unit economics (2023–2026), with MindChamps (SGX) as the public proxy. A lens on economic quality, not a verdict on an owner-operated livelihood; several SG cost and startup figures are proxies or folk figures, flagged where used.
Model your own centre — the whole game is occupancy against the fee cap:
Break-even occupancy
51%
of capacity filled, just to break even (~41 children)
Monthly net
S$15,040
at 75% full
Manpower / fees
29%
healthy
Verdict: Healthy — but pressure-test occupancy; in a falling-birth market it is the fragile input, not the fee.
Illustrative model on ECDA ratios (Code of Practice, 2025) + a MindChamps-derived ~67% manpower anchor (FY2025 SGX). Set the fee to ~S$610 (AOP) or ~S$650 (POP) to feel the cap; ~S$1,200–2,500 for private/premium. SG per-centre margins are not cleanly published — a starting frame, not financial advice. Manpower gauge: ≤60% healthy, 60–72% tight, >72% danger.
How to actually open one (in the order that matters)
Almost every guide leads with the ECDA licence application. That's the last step. The real work — and the real timeline — is the premises, the fit-out and the educators, all of which must be ready before ECDA visits.
The approval stack, sequenced
- ACRA — register the operating entity (company, society, charity or co-op).
- Secure premises — most centre space is government-gated (HDB void-deck / state sites by tender) or URA-approved commercial. Plan for change-of-use approval. Gate 1 — the real bottleneck.
- Fit-out to ECDA standards — minimum floor area per child (more for infants), safety, hygiene and an outdoor / gross-motor play area; plus SCDF fire safety and the other premises clearances. Gate 2.
- Hire a qualified Principal + certified educators — ECDA-certified, in the right numbers for your ratios (infant 1:5, and higher for older groups), into a market with a real teacher shortage. Gate 3 — the hardest to solve.
- ECDA licence application via the GoBusiness Licensing Portal — the premises must pass a site visit; the licence is then issued in ~14 working days (initially 12 months).
- Optional but strategic: apply for a Partner Operator (POP) appointment for government funding — but it is a competitive 5-year tender with limited slots, and it brings the fee cap with it.
The decision most founders get backwards
Before any of this, decide your tier, because it sets every other number. Funded & capped (AOP/POP) gives you recurrent government support and a near-free fee for parents — volume and stability, but a capped ceiling and a thin, funding-dependent margin. Private / premium (uncapped) gives you pricing freedom and the only real margin — but you must earn a S$1,500–2,500 fee with a credible curriculum, location and reputation, and you carry full occupancy risk in a softening market. There is no good “middle” — an uncapped centre priced like a capped one just loses money slowly. Pick the end you can win, and build the whole plan around it. And verify the current rules against the ECDA Code of Practice (2026 edition) — the ratios, floor-area and tenure rules are updated periodically.
Where a new centre actually wins
“Saturation” is the wrong frame and “another full-day centre on price” is a losing one. With a national surplus of capped places and a shrinking cohort, the winners are not generic — they own a segment the funded giants under-serve, or a quality story that escapes the cap.
Infant care (if you can staff it)
The least-saturated centre-based segment — demand outruns supply, but the binding constraint is qualified Infant Educators at a punishing 1:5 ratio, not buildings. Solve the staffing and you have the tightest part of the market.
Early intervention & inclusion
A genuine demand-exceeds-supply pocket — EIPIC waits run months, and ECDA is pushing inclusive places (InSP) into mainstream preschools. Specialist-manpower-gated and government-shaped, but a long, under-served runway.
Premium school-readiness / bilingual
The clearest way to escape the fee cap: an uncapped centre with a credible Primary-1 readiness and Chinese/bilingual story that anxious dual-income parents will pay S$1,500–2,500+ for.
The specific outgrowing estate
National slack hides pockets of tightness in fast-growing BTO towns (Tengah, Punggol, Tampines North). Target the estate outgrowing supply — but verify per-estate utilisation, because the national signal is a surplus.
Flexible & extended hours
The 5-day-week reform and the childminding pilot both signal unmet demand for drop-in, half-day and shift-worker care — a thin, underserved niche the mass-market giants ignore.
The honest AI/tech edge
The certain wins are boring: automated enrolment, waitlist and parent-comms (occupancy is the whole game), plus rostering to the ECDA ratios and admin that frees educators to teach. Treat “AI curriculum” as marketing, not a moat — the real lever is filling and keeping the places.
Questions founders ask
How much does it cost to open a childcare centre in Singapore?
There is no official figure, and the widely-quoted "~S$800,000+" is a consultancy estimate, not a government number — treat it as a folk figure. What is certain: the ECDA licence itself is trivial (a fee under the ECDC Regulations, immaterial next to the real cost), while the binding costs are the premises (most centres sit in HDB void-deck or commercial space tendered through the government), the fit-out to meet ECDA floor-area, safety and outdoor-play rules, and several months of manpower before you fill the places. Budget for a large six-figure sum, dominated by renovation, deposit and pre-opening payroll — and verify every line, because Singapore publishes no clean per-centre startup cost.
What licence do I need to open a childcare centre in Singapore?
A licence from ECDA (the Early Childhood Development Agency) under the Early Childhood Development Centres Act, applied for through the GoBusiness Licensing Portal. ECDA licenses by class: Class A (infant care, 2–18 months), Class B (childcare, 18 months to below 7), and Class C (kindergarten) — you can hold A, B, C or a combination. A new licence runs 12 months; on renewal ECDA assigns a 6, 12, 24 or 36-month tenure based on your track record (36 months is the reward tier, only after two consecutive 24-month licences). The premises must be ready and pass a licensing site visit, with the licence issued within about 14 working days of that visit. Note: an ECDA Class C kindergarten is not the same as an MOE Kindergarten — those are run directly by MOE and you cannot open one. Verify the current rules against the ECDA Code of Practice (2026 edition).
What is the childcare fee cap in Singapore for 2026 (Anchor Operator vs Partner Operator)?
From 1 January 2026, the monthly full-day childcare fee caps (before GST) are S$610 at Anchor Operator (AOP) centres and S$650 at Partner Operator (POP) centres; full-day infant care is capped at S$1,235 (AOP) and S$1,290 (POP). These caps were cut by S$30 from the 2025 levels and have been ratcheted down repeatedly — any guide quoting "S$720 / S$800" is stale. AOPs (PCF Sparkletots, NTUC My First Skool, M.Y World, Skool4Kidz, E-Bridge) and POPs accept government funding in exchange for keeping to these caps and meeting quality and educator-pay requirements. Private and premium centres are not capped and commonly charge S$1,500–2,500+ a month. (Source: ECDA / MSF, effective 1 Jan 2026.)
Is a childcare or preschool business profitable in Singapore?
It can be, but margins are thin and the whole game is occupancy. The only listed pure-play proxy, MindChamps PreSchool (SGX), ran roughly 3.8% net margin in FY2025 — and even that leaned on one-off disposal gains, with the underlying operating business around breakeven. Manpower (teacher salaries) is the dominant cost at roughly two-thirds of revenue, and ECDA keeps raising salary floors, so a capped AOP/POP centre needs government funding just to clear breakeven. Singapore publishes no clean per-centre net margin. The realistic read: the affordable tier survives on subsidy and scale; the room for a real margin is in premium, well-occupied, differentiated centres — where occupancy and educator retention, not the fee, decide whether you make money.
Want a report like this for your company or sector?
Tell us what you're trying to figure out. We'll build you a Deep-Context report — the same depth, focused on your business.
About this report. Built with SGAI's Deep-Context Engine — human-directed, AI-accelerated. Figures draw on ECDA (the Code of Practice, the AOP/POP fee caps, the subsidy tables and salary ranges, and the data.gov.sg centre listing), MSF (Committee of Supply 2026 and the end-2024 places update), MOE, SingStat / population.gov.sg, MOM, and MindChamps PreSchool's SGX filings (2024–2026). Singapore publishes no clean per-centre unit economics or startup cost, so those figures are proxies (MindChamps' audited cost ratios) or flagged folk estimates — with the caveats stated inline. The MOAT Score is a transparent SGAI judgement on economic quality, not a verdict on an owner-operated livelihood. Verify fees, ratios and regulatory steps against ECDA and the 2026 Code of Practice before acting; this is not financial, legal or regulatory advice.
We're not a grant office — we build AI and intelligence for Singapore SMEs. Talk to us.