Starting an online store in Singapore, decoded.
Almost zero barrier to open — and almost no moat once you do. Anyone can list on Shopee by lunchtime, which is exactly why a generic store competes on one thing only: price, to the floor. The platform's cut keeps rising, ads eat the rest, and the “US$9 billion market” you read about is gross merchandise value — not a single dollar you keep. Here is the picture the dropshipping gurus skip: the real numbers, the platform treadmill, the GST and ACRA rules, and the one route that actually builds a business.
~14%
of SG retail is online (Feb 2026)
~2% → ~12%
Shopee's take-rate, 2018 → 2026
~47%
of sellers: lifetime profit < US$25k
S$115
to register (the easy part)
The MOAT Score: is an online store 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. An online store scores a hair above a café: asset-light enough that you can fail cheaply, but a no-moat commodity where the platform and the ad networks capture the value.
Winner-takes-most
Winner-takes-most — hard.
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
7/25Does the average operator actually keep money — real net margin and return on the capital tied up?
This sector: thin-to-negative for a generic reseller once COGS, the platform take, ads, payment and returns are allocated; ~47%Global or regional figure — not Singapore-specificJungle Scout, State of the Amazon Seller, 2024US/global survey of ~2,000 sellers — illustrative of the structural economics, not Singapore-specific. of sellers have lifetime profit under US$25k.
Buffett 1979 — “a high earnings rate on equity capital… without undue leverage”; 1986 owner earnings.
Operating moat
5/25Pricing power and a durable competitive advantage — can a typical operator raise prices and have customers shrug?
This sector: a price-taker with no moat — identical listings, Temu/Shein/Taobao set the floor, and the platform owns the customer, not you.
Buffett, FCIC 2010 — pricing power is “the single most important decision”; 1991 franchise; 2007 moat.
Appetite
14/25Demand durability — steady, recession-resistant repeat demand vs fragile, discretionary or faddish.
This sector: online demand is genuinely broad and growing (70.6% of SG adults bought online in 2024) — but discretionary, price-driven, and a single store’s slice is fragile.
Graham, Security Analysis Ch.2 — inherent stability “derives from the character of the business”.
Treadmillinverted · less is better
9/25Capital intensity and structural drag — rent, churn, fashion, discounting. Scored inverted: less treadmill, more points.
This sector: asset-light is the one real edge (start for ~S$3–15k), but rising platform take-rates, ad-cost inflation, returns and platform-wipeout risk (Qoo10) make it a brutal operating treadmill.
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: the thing with no barrier to entry has no barrier to competition
An online store is the easiest business in Singapore to start — and that is precisely the problem. You can register a sole proprietorship for S$115, list on Shopee the same afternoon, and be “open” with no lease and no fit-out. But every barrier that's low for you is low for the next ten thousand people. With identical products sourced from the same suppliers, the only lever left is price — and there is always someone willing to make less than you. Buffett's word for this is commodity: undifferentiated products, infinite supply, returns competed down to the cost of capital. The only durable winner in a commodity is the lowest-cost operator at scale, and against Temu, Shein and Taobao shipping factory-direct into Singapore, that is not going to be you.
Here is the part the “quit your job and dropship” videos never show: you don't even own the customer. On a marketplace, the buyer belongs to Shopee, not to you. You rent access to them — and the rent keeps going up. Shopee's effective take-rate has climbed from around 2% in 2018 to roughly 12% by 2026; in 2024 alone it raised commissions “by about a third,” twice, and in February 2026 it added a new 5% cross-border fee. The incentive is obvious — once sellers are locked in, the platform captures more of every sale. Munger's rule: show me the incentive and I'll show you the outcome.
And the economics are brutal even for the house. Sea, Shopee's parent, was founded in 2009 and didn't post its first quarterly profit until Q4 2022 — and it got there by hiking seller fees and cutting subsidies. If the platform needed 13 years and your fees to make money, ask what that leaves for the generic tenant. Worse, platform dependence can erase you overnight: when Singapore-based Qoo10 was wound up as insolvent in November 2024, sellers' unpaid payouts — their money, held by the platform — ran to roughly S$900 million. Your cash was its working capital, until it wasn't.
Why a generic store can't win on price
A reseller is fenced in from below: factory-direct cross-border sellers and identical listings set the floor, so a generic store can only match it — while the platform's cut comes off the top either way.
- Temu / Shein / Taobao (factory-direct)S$12
the price floor a reseller cannot beat
- Generic marketplace reselleranchorS$18
the same product, a few dollars more — racing to the bottom
- Niche / curated own-listingS$32
- Real own-brand / private labelS$55
a brand customers ask for by name — the only escape
Source: SGAI illustrative price ladder for a comparable consumer good, 2026
The map: online is ~14% of SG retail — and the “US$X billion market” is GMV, not your revenue
Start with the only honest, Singapore-specific number. Per SingStat's monthly Retail Sales Index, 14.1%SourceSingStat Retail Sales Index, Feb 2026 of total retail sales were online (~16% excluding motor vehicles) — a settled mid-teens share that peaked at 24.9%SourceSingStat / data.gov.sg, Apr 2020 during COVID and then fell back. The credible market figure is ~US$9BSourcee-Conomy SEA (Google/Temasek/Bain), 2024Explicitly GMV — gross merchandise value flowing through platforms, not seller revenue. ~6% of SEA's US$159B. of GMV — about 6% of Southeast Asia. The bigger figures you'll see — up to US$116BShaky figure — treat with cautionIMARC GroupPublished 'Singapore e-commerce' estimates range from ~US$5.6B to ~US$116B — a ~20x spread — because of GMV-vs-revenue confusion, scope creep (B2B, travel, food) and SEA-vs-SG mix-ups. Anchor on SingStat + e-Conomy GMV. — disagree by up to twenty times because they confuse GMV with revenue, bundle in B2B and travel, or quote the whole region. The trap is reading any of them as your opportunity.
~14%
of SG retail is online
SingStat RSI, Feb 2026 (settled, not surging)
~US$9B
SG e-commerce GMV
e-Conomy SEA 2024 — GMV, not revenue
56% online
computers & telecom
vs supermarkets only ~11% (SingStat)
~20x
spread in 'market size' estimates
US$5.6B to US$116B — mostly junk
Online's share of SG retail spiked, then settled
The number the hype cycle remembers is the COVID peak. The honest read is the line after it: online retail settled into the mid-teens, not an exploding frontier.
Apr 2020—COVID peak — the figure the hype still quotes
Feb 2026—settled mid-teens (SingStat RSI, total basis)
Stale: the 25% peak older guides extrapolate as the trend
Source: SingStat Retail Sales Index, online share of total retail (Feb 2026 = 14.1%); Apr 2020 peak 24.9% (data.gov.sg). Intermediate points indicative of the settle; 2019 pre-pandemic ~6%.
The players: the marketplaces you rent, and the brands that escaped them
The single most predictive question isn't which platform you sell on — it's whether you own anything: a brand, a product, a customer relationship. The marketplaces are channels you rent at a rising price. The Singapore success stories almost all did the opposite — they built their own product and owned the customer, then used marketplaces as a channel, not the business.
Rent the channel, or own the brand?
| Model / player | What it is | Owns the customer? | Pricing power? | Signal |
|---|---|---|---|---|
| Shopee (Sea)#1 in SG by GMV (~52% of SEA) | Marketplace you rent — take-rate ~2%→~12% | First profit only in Q4 2022, after 13 years | ||
| Lazada (Alibaba) · TikTok Shop · Amazon.sgThe other rented channels | Marketplaces; fees rising (Lazada +3%, Feb 2026) | TikTok Shop SG actually shrank >50% YoY in H1 2025 | ||
| Qoo10The cautionary tale | SG-based marketplace, over-leveraged | Wound up insolvent Nov 2024; ~S$900M seller payouts frozen | ||
| SecretlabOwn-brand gaming chairs (2014) | Proprietary design, DTC, global | 500k+ chairs/yr, 60+ countries; bootstrapped ~S$50k | ||
| Love, Bonito · Castlery · Allies of SkinOwn-label DTC that scaled | Own product + owned customer + repeat | Started small (Allies of Skin: a Shopify store, hand-packed) |
Shopee (Sea)
#1 in SG by GMV (~52% of SEA)
- What it is
- Marketplace you rent — take-rate ~2%→~12%
- Owns the customer?
- Pricing power?
- Signal
- First profit only in Q4 2022, after 13 years
Lazada (Alibaba) · TikTok Shop · Amazon.sg
The other rented channels
- What it is
- Marketplaces; fees rising (Lazada +3%, Feb 2026)
- Owns the customer?
- Pricing power?
- Signal
- TikTok Shop SG actually shrank >50% YoY in H1 2025
Qoo10
The cautionary tale
- What it is
- SG-based marketplace, over-leveraged
- Owns the customer?
- Pricing power?
- Signal
- Wound up insolvent Nov 2024; ~S$900M seller payouts frozen
Secretlab
Own-brand gaming chairs (2014)
- What it is
- Proprietary design, DTC, global
- Owns the customer?
- Pricing power?
- Signal
- 500k+ chairs/yr, 60+ countries; bootstrapped ~S$50k
Love, Bonito · Castlery · Allies of Skin
Own-label DTC that scaled
- What it is
- Own product + owned customer + repeat
- Owns the customer?
- Pricing power?
- Signal
- Started small (Allies of Skin: a Shopify store, hand-packed)
The cautionary tale: Naiise
A celebrated Singaporean retailer that curated and resold other people's products across heavy physical and online footprint — and whose founder filed for bankruptcy after the business was liquidated in 2021, with payment troubles dating back years. The lesson is the thesis of this whole report: footprint and traffic without a defensible product or healthy unit economics is a treadmill, not a moat. Reselling other people's goods — online or off — doesn't become a business just because it gets big.
The customer: a nation of online shoppers — and the world's most ruthless deal-hunters
The demand is real: 70.6%SourceIMDA via SingStat / data.gov.sg, 2024 of Singapore residents aged 18+ bought something online in 2024 — rising to 92.9% of those aged 18–39. But where they buy and how they decide is the catch. Two in three reach for a marketplace app; only about three in ten use a specific retail brand's app — so “default” means Shopee, not you. And they are relentless on price: 93%SourceAirwallex × Statista, 2025Commissioned survey, n=1,000 SG cross-border shoppers — sample skews to savvy/cross-border buyers; flag. plan purchases around major sale events, and 69%SourceAirwallex × Statista, 2025 buy from overseas at least monthly — most often from China. The Singaporean shopper is structurally disloyal to any seller who competes only on price.
How Singaporeans reach an online store
Channel preference for SG online shoppers (indicative shares; people use more than one). The marketplace app is the front door — which is exactly why a generic store has no direct relationship to defend.
- Marketplace apps (Shopee / Lazada / Amazon)62%YouGov 2025: ~67% use marketplace apps — the most-used channel; the customer is the platform’s
- Brand / retailer own apps & sites27%YouGov 2025: ~29% — the channel where a real brand owns the relationship
- Social / live commerce (TikTok / IG)11%present but SG is the SEA laggard — TikTok Shop SG shrank >50% YoY in H1 2025
Source: YouGov 2025 (marketplace vs brand-app share, SG) + IMDA/SingStat 2024 (online-shopper penetration). Shares are channel-usage, not GMV, and overlap — indicative. Live-commerce share is directional.
The economics: GMV is vanity; what you keep is the whole game
Starting is cheap — ~S$3–15k covers registration, stock and a few months of ads for most first stores. That low capital at risk is the genuine upside of e-commerce: you can fail without a six-figure lease hanging over you. The danger is the per-order maths. A generic reseller's dollar is eaten from three directions at once: the platform take (~10–16% all-in and rising), ads to be seen at all (often 15–25% of revenue, climbing as Meta's ad prices rise ~9–14% a year), and returns — before you even count your cost of goods. A 65% gross margin can be a 5% net margin, or less, once everything is allocated. The platform-fee inputs below are Singapore-specific; the margin and ad-cost benchmarks are global, and flagged as such.
All-in to start (and survive the first few months)
ACRA registration, initial stock, packaging, and a few months of ads/testing for a typical first store. The real money isn't the setup — it's the working capital to keep buying ads and inventory while you find out whether anyone wants what you sell. Asset-light is the upside; cheap to start means cheap to fail.
Source: SGAI synthesis of SG operator guidance, 2026
Where the online-store dollar goes (generic reseller)
A typical marketplace reseller, modelled per S$100 of sales. Three parties — your supplier, the platform, and the ad network — take their cut before you do. What's left is thin to negative.
- COGS (landed cost of goods)−60%40% left
A reseller buys wholesale; private label can push this far lower · Global reseller benchmark
- Platform take (commission + transaction + programmes)−12%28% left
SG all-in ~10–16% and rising — Shopee added a 5% cross-border fee Feb 2026 · SG platform fee schedules 2026
- Ads / customer acquisition−15%13% left
You must buy ads to be seen; Meta ad prices rise ~9–14%/yr · Meta 8-K + global CAC
- Payment + shipping + packaging−6%7% left
- Returns & refunds drag−4%3% left
Online returns ~17% (apparel far higher) — a margin-killer sellers under-budget · NRF 2024 (global)
What the owner actually keeps
Verdict: A thin 3% — survivable, not comfortable; one bad assumption flips it negative.
Illustrative model on SG platform fees (2026) + global margin/CAC benchmarks; e-commerce net margins for generic resellers are thin to often negative. A real own-brand with repeat customers changes this picture entirely. Not financial advice.
E-commerce / online store
Would a value investor own the average operator here?
A value investor would not want the average online store — a generic reseller is a price-taker with no moat, on a platform-and-ad-cost treadmill, who doesn't even own the customer. The asset-light economics mean you fail cheaply, not that you win.
Identical listings and factory-direct cross-border sellers set the floor; a generic store can only match it.
Buffett, FCIC 2010 — pricing power is “the single most important decision”
Products copy in hours, competitors are infinite, and the platform owns the customer — what little edge exists erodes as take-rates rise.
Buffett 2007 — an enduring moat protects returns on capital
Capital is light, but so are the returns on it for a reseller — often negative after ads + returns.
Buffett 1979 — a high earnings rate on capital, unleveraged
The one genuine advantage: no lease or fit-out — start for ~S$3–15k and fail cheaply.
Buffett 2007 — the worst business needs much capital; this one needs little
Online demand is broad and growing, but discretionary, price-driven, and not loyal to any one seller.
Graham, Security Analysis Ch.2 — inherent stability is qualitative
Take-rates rise once you’re locked in (Shopee ~2%→~12%); ad costs climb ~9–14%/yr; payouts can be frozen (Qoo10).
Bloomberg 2024 + Meta 8-K + ST (Qoo10) — the rent keeps rising
A real own brand or private label, an owned customer relationship (email/community/first-party data) and repeat purchase — not a commodity anyone can list. Treat marketplaces as a channel, never the business.
Assessment uses the value-investing lens on the AVERAGE operator — a generic marketplace reseller — using SG platform fees (2026) and global margin/CAC benchmarks. A lens on economic quality, not a verdict on an owner-operated side business.
Model your own per-order maths — drag the sliders:
You keep, per order
S$1
a 2% net margin on a S$40 order
Monthly net
S$190
on S$12,000 sales
Leaks to others
29%
crushing
Verdict: Thin — survivable only at volume with no surprises. One fee hike or a returns spike flips it negative.
Illustrative model on platform-fee + global margin benchmarks (2024–2026). “Leaks to others” = platform + ads + payment, before your own product cost — this is why GMV is not what you keep. A starting frame, not financial advice.
How to actually start one (the rules that catch sellers out)
The setup is the easy part. The traps are tax and product rules — the questions everyone Googles at 1am, and where the official answers are precise.
Register & comply, in order
- ACRA — register if you sell for profit on an ongoing basis (sole prop ~S$115; Pte Ltd S$315). Selling strictly under your own NRIC name is the exemption.
- Home-Based Business Scheme — no HDB/URA licence needed to run a small online store from home; but no non-resident staff, no shopfront/signage, no bulk storage.
- Income tax (IRAS) — profits are taxable if it's a trade; file if net trade income > S$6,000 or total income > S$22,000. Selling your own used items at no profit is not taxable.
- GST (IRAS) — register and charge 9% only above S$1,000,000 taxable turnover. Below that it's voluntary.
- Product licences — cosmetics need an HSA notification; food import needs SFA; alcohol needs a liquor licence; vapes are banned outright.
- PDPA + Spam Control — consent before collecting data, a mandatory DPO (no SME exemption), DNC checks before marketing, and unsubscribe/“<ADV>” on promotional messages.
The detail nobody tells you: the cross-border GST loophole is closed
Since 1 January 2023, imported low-value goods of S$400 or less attract 9% GST at checkout (collected by GST-registered overseas vendors). The old “cheap overseas = no GST” arbitrage that local sellers used to lose to is gone — a small but real levelling of the field, and a fact most guides still get wrong. The bigger competitive reality, though, is unchanged: Singaporeans are among the world's most active cross-border shoppers, so a local reseller is competing with the entire planet's catalogue, GST-paid.
Where a new online store actually wins
Reselling a commodity won't survive the take-rate treadmill. Every Singapore success did the same thing: built something of their own, and owned the customer. The escape is always away from “anyone can list this” toward “they ask for it by name.”
Own brand / private label
The single highest-leverage move: a proprietary product lifts gross margin from ~25–35% (reseller) toward ~65%, and it can’t be undercut by an identical listing. This is what every SG winner did — Secretlab, Love, Bonito, Allies of Skin.
Own the customer, not the channel
Capture email, build a community, sell off-platform. With ad costs rising ~9–14%/yr, repeat purchase is the only maths that works — and a relationship is the one asset the platform can’t take.
A defensible niche
Don’t sell what everyone sells. A focused niche with genuine expertise, curation or craft (Bynd Artisan) earns pricing power a generic catalogue never will.
Subscription & repeat
Consumables, refills, and subscriptions turn one-off, ad-bought buyers into recurring revenue — the closest an online store gets to a moat.
Marketplaces as a channel, not the business
Use Shopee/Lazada to acquire and test, then move repeat customers to your own site to keep the margin and the data. Never let one platform be your whole business — ask Qoo10’s sellers.
The honest AI edge
AI genuinely helps: multilingual listings, customer-service chat, demand forecasting, ad creative. But it lowers cost identically for every competitor, so it accelerates the race to the bottom for a reseller. It compounds a real brand and commoditises a generic one. AI sharpens the knife; it doesn’t give you a knife.
Questions founders ask
Do I need to register my online business with ACRA in Singapore?
Yes, if you sell for profit on an ongoing basis — including on Shopee, Lazada, Amazon.sg, TikTok Shop, Carousell or Instagram — unless you trade strictly under your own full NRIC name. A sole proprietorship costs about S$115 to start (S$15 name + S$100 registration) and ~S$30/year to renew; a private limited company costs S$315 to incorporate. (ACRA, 2026.)
Do I need to pay income tax or GST on my Shopee / online-selling income?
Income tax: yes, if it is a trade (you source and sell regularly to make a profit), even as a side hustle — declared under "Trade, Business, Profession or Vocation"; selling your own used personal items at no profit is not taxable. You must file if net trade income exceeds S$6,000 or total income exceeds S$22,000 a year. GST: only if your taxable turnover exceeds S$1,000,000 a year (then you charge 9% GST). Separately, since 1 Jan 2023 imported low-value goods of S$400 or less now attract 9% GST at checkout, so the old "cheap overseas = no GST" advantage is gone. (IRAS, 2026.)
Is selling on Shopee or Lazada actually profitable in Singapore?
For a generic reseller, rarely. After the platform take (roughly 10–16% all-in and rising), ads to be seen, payment fees, shipping and returns, plus your cost of goods, the average reseller keeps a thin-to-negative margin. Globally, about 47% of Amazon sellers have lifetime profits under US$25,000 and roughly 2% of sellers make over half of all seller revenue. Even Shopee’s parent, Sea, took 13 years and repeated fee hikes to post its first profit. The sellers who make money almost always sell an own brand or private label, not a commodity anyone can list.
Shopee vs Lazada vs TikTok Shop — which has the lowest seller fees?
Rough all-in (category commission plus the ~2% transaction fee, for an established Singapore seller, early 2026): Lazada ~9–10%, TikTok Shop ~6–10%, and Shopee ~10–12% — the highest after its January 2026 hike and a new 5% cross-border technical-support fee. TikTok Shop folds payment processing into its commission; Shopee and Lazada charge it separately. These all-in totals are estimates, not official disclosures, and fees changed in early 2026 — verify in each platform’s seller centre before deciding.
Do I need a business licence to run an online store from home?
No HDB or URA licence is needed for a compliant small-scale home-based online business under the Home-Based Business Scheme — but no non-resident staff, no physical shopfront or signage at your home, and no bulk storage or nuisance. You still need ACRA registration, and a product-specific licence if your goods are regulated: cosmetics need an HSA product notification, food import needs SFA approval, alcohol needs a liquor licence, and vapes are banned outright. (URA / HDB / GoBusiness, 2026.)
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 SingStat, IMDA, IRAS, ACRA, HSA, the e-Conomy SEA report (Google/Temasek/Bain), company filings and platform fee schedules (2023–2026). Singapore-specific data (online retail share, shopper penetration, GST/tax rules, platform fees) is separated from global benchmarks (margins, CAC, return rates), which are flagged as illustrative. The MOAT Score judges the average operator — a generic marketplace reseller — on economic quality; it is not a verdict on an owner-operated side business. Platform fees change often; verify each seller centre and each agency before acting.
We're not a grant office — we build AI and intelligence for Singapore SMEs. Talk to us.