Last updated 31 August 2026 — by Adrian Tan, Singapore Digital Marketing
Nine days from now, Singapore’s online retail machine starts its four-month run: 9.9, then 10.10, then 11.11, then 12.12 and Christmas. Almost every guide written about it treats those four dates as equivalent, tells you to “prepare early”, and moves on.
They are not equivalent. Singapore’s own official retail statistics show a clear and repeated shape across 2024 and 2025, and it contradicts the usual advice in two specific ways: September matters more than October, and December is not the online peak — November is. If you are allocating a finite budget and a finite team across four sale events, that changes what you do.
This guide starts with that data, then covers the three rulebooks that actually bite on a Singapore sale — CCCS on how you may present a discount, IRAS on how you may display a price, and the platforms’ own bidding mechanics — and finishes with a countdown you can run against.
What the official data actually shows
The Department of Statistics publishes the online share of retail sales monthly. Here is that series for retail trade excluding motor vehicles, and for the two categories where the mega-sales bite hardest, for the last two complete years.
| Online share of sales | Aug | Sep (9.9) | Oct (10.10) | Nov (11.11) | Dec (12.12) |
|---|---|---|---|---|---|
| All retail excl. motor, 2025 | 15.2% | 17.6% | 16.8% | 19.3% | 17.0% |
| All retail excl. motor, 2024 | 14.3% | 16.2% | 14.5% | 16.7% | 15.3% |
| Computer & telecoms equipment, 2025 | 56.6% | 56.9% | 52.0% | 60.6% | 56.5% |
| Furniture & household equipment, 2025 | 33.3% | 35.8% | 36.5% | 40.8% | 33.4% |
| Supermarkets & hypermarkets, 2025 | 11.3% | 12.2% | 12.5% | 12.7% | 11.7% |
Four things fall out of that, and each one has a budget consequence.
November is the peak, in both years, by a clear margin. The online share of all retail excluding motor vehicles hits 19.3% in November 2025 — the highest month in the series — against 17.0% in December. The same ordering holds in 2024 (16.7% against 15.3%) and in 2023 (17.0% against 14.5%). If you have one month of real budget, it is November, not December.
September steps up hard, and October falls back. August to September 2025 moves 15.2% to 17.6%. October then drops to 16.8%. In 2024 the drop is sharper: 16.2% in September down to 14.5% in October. On this evidence, 10.10 is the weakest of the four dates in Singapore, and September is the second-strongest month of the year. That is an argument for treating 9.9 as a real event rather than a warm-up — which is precisely why this is worth reading in late August.
The effect is category-specific and enormous where it lands. Computer and telecommunications equipment runs at 60.6% online in November 2025, up 8.6 points from October. Furniture and household equipment moves from 36.5% to 40.8%. Supermarkets barely move at all: 12.5% to 12.7%. If you sell groceries or everyday consumables, the mega-sale calendar is largely somebody else’s event and you should plan accordingly rather than burning budget into an inflated auction.
December is a physical-retail month. The online share drops in December in all three years even as total retail rises. Christmas gifting in Singapore still happens substantially in stores and malls. So 12.12 is not the crescendo the name implies; it is a smaller online event sitting inside a large offline one.
One honest caveat, because it matters: SingStat reports the online share of retail sales by month. It does not attribute sales to specific sale days, and it does not separate a 9.9 promotion from ordinary September trading. The monthly shape is strong and repeats, but treat it as evidence about months, not proof about individual dates.
Rulebook one: CCCS decides how you may present a discount
This is the part most sale planning skips, and it is enforceable. The Competition and Consumer Commission of Singapore’s Guidelines on Price Transparency came into effect on 1 November 2020 under the Consumer Protection (Fair Trading) Act, and they apply to all suppliers, online and physical. They name four practices. Every one of them is something a mega-sale campaign does by default.
Drip pricing
CCCS expects unavoidable or mandatory charges — taxes, surcharges, fees — to be incorporated into the headline price. Where a charge genuinely cannot be calculated in advance, its existence must be disclosed clearly and prominently alongside the headline price. Optional add-ons should be opt-in or opt-neutral, not pre-selected.
In practice, on an 11.11 campaign: a “$29” hero price that becomes $34.90 at checkout after a mandatory handling fee is the exact pattern the guidelines address. Pre-ticked insurance, pre-ticked express shipping and pre-ticked gift wrap all fall on the wrong side of the opt-in expectation.
Price comparisons
Comparisons with a competitor’s price must not be false or misleading, and must compare genuinely similar or equivalent products. CCCS expects suppliers to conduct regular research and to keep records substantiating reference prices.
Discounts — the strike-through problem
This is the big one. Where you offer a discount or compare against your own usual or previous price — a strike-through, in other words — the price benefit must be genuine and you must have a valid basis for the comparison. Time-limited discounts must state the period clearly and accurately. CCCS encourages suppliers to keep records of past sales and prices to prove the past price and the discount were real.
The practical failure mode in a Singapore mega-sale is well known: raise the list price in late October, strike it through on 11 November, and call it 40% off. If the higher price was never a price at which you meaningfully sold, the comparison has no valid basis. CCCS has publicly required traders to cease misleading sale advertising before — the “Closing Down Sale” and “Fire Sale” case, and the Expedia matter on the validity period of “Daily Deals”, are both on the public record.
The operational instruction that follows is simple and almost nobody does it: before you set any strike-through price, export your actual selling prices for the preceding weeks and keep the file. If you cannot evidence the “was” price, do not publish it. Use “$X off” or a bundle instead — both are unambiguous and need no historical claim.
The word “free”
“$0” and “free” claims must be accurate, with qualifiers, subsequent or deferred charges and key terms stated prominently alongside. Free-trial users must be notified before charges begin and told how to cancel. “Free shipping” with an unstated minimum-spend threshold is the version of this that appears in nearly every sale campaign.
Rulebook two: IRAS decides how you may display the price
Separate law, separate regulator, routinely broken during sales.
GST-registered businesses must display and quote GST-inclusive prices on all price displays to the public — price tags, price lists, advertisements, brochures and your website. If both GST-inclusive and GST-exclusive prices appear, the inclusive price must be at least as prominent. Failure to comply carries a fine of up to $5,000.
And the one that catches sale campaigns specifically: if you are GST-registered and intend to give a discount equal to the GST amount, you should not advertise “no GST”. IRAS is explicit that this is misleading, because GST is still included in the post-discount price and you still have to account for it. “GST absorbed” and “we pay your GST” creative needs rewriting as a straightforward discount — “9% off”, which is what it is.
There is a narrow exception for hotels and F&B establishments that impose a service charge; they must still display a prominent statement that prices are subject to GST and service charge.
If you sell into Singapore from overseas, one more rule applies. Since the low-value goods regime, GST is payable at the point of purchase on goods with a sales value of S$400 or below, imported by air or post, and bought from GST-registered suppliers — including via a GST-registered marketplace. Above S$400, or by sea or land, GST is collected at importation instead. If a customer discovers the tax after checkout, that is both a conversion problem and a drip-pricing problem at the same time. Our guide to payment gateways in Singapore covers the checkout side of this in more depth.
Rulebook three: the ad platforms’ own mechanics
The two most expensive mistakes in a Singapore mega-sale are both self-inflicted, and both are documented by the platforms themselves.
Meta: do not reset the learning phase on 10 November
Meta’s delivery system puts an ad set into the learning phase whenever it is created or significantly edited, and the ad set exits after about 50 results in the week following its last significant edit. Meta lists budget changes among the edits that can cause an ad set to re-enter learning, and advises setting a budget large enough to get enough total results while avoiding frequent budget changes.
Now consider what almost every advertiser does the day before 11.11: triples the budget. That is a significant edit. The ad set re-enters learning at the exact moment performance matters most, delivery becomes unstable, and CPA rises — and because the sale lasts a day or two, it never accumulates the 50 results needed to stabilise before the event is over.
The fix is structural, not clever. Scale the budget in steps in the weeks before, so the ad set is already exercising at something near sale-day spend and is out of learning when the day arrives. Where a genuinely different offer needs a genuinely different ad set, launch it early enough to exit learning. Our note on scaling a Meta campaign without breaking it covers the increments; what Meta ads cost in Singapore covers the baseline you are scaling from.
Google: use seasonality adjustments, and use them correctly
Google Ads provides seasonality adjustments for exactly this situation: telling Smart Bidding to expect a conversion-rate change for a future event. Google’s guidance is precise and worth following to the letter.
- They are ideal for short events of 1 to 7 days, and Google says they may not work as well over extended periods of more than 14 days.
- Use them only if you expect major conversion-rate changes, because Smart Bidding already handles ordinary seasonality.
- You are scheduling a conversion-rate adjustment: if you expect conversion rate to rise 50% during a three-day sale, apply up to a +50% adjustment.
- They are available for Search, Shopping and Display on Target ROAS and Target CPA, and for Performance Max and App campaigns on all bid strategies.
Two practical consequences. First, 11.11 is a textbook fit — a one-to-three-day event with a large expected conversion-rate lift. A month-long “November Sale” is not, and applying an adjustment across it will do more harm than good. Second, be honest about the number. A +100% adjustment on a sale that delivers +20% teaches the bidder the wrong thing and the correction lands after the event. If you have last year’s data, base the figure on it. If you do not, be conservative. More on the underlying mechanics in how Smart Bidding actually works and, for product feeds, Google Shopping in Singapore.
The countdown
Working backwards from a sale date. The dates below are written for 11.11; shift them for 9.9 or 12.12.
Eight weeks out — fix measurement, then decide the offer. If conversion tracking is broken, everything downstream is guesswork and the bidding algorithms will be optimising to noise on the highest-spend day of your year. Confirm the offer and check the margin at the discounted price, including payment processing and any returns you expect.
Four weeks out — creative, and the evidence file. Build the assets. Then do the unglamorous thing: export your actual selling prices for the preceding weeks and archive the file. That single step is what makes a strike-through defensible under the CCCS guidelines, and it takes ten minutes.
Two weeks out — start scaling, not on the day. Step budgets up so ad sets stabilise. If you are launching new creative, launch it now so it has accumulated results before the event.
Three days out — adjustments in, edits frozen. Set the Google seasonality adjustment for the sale window only. Stop editing Meta ad sets. Warm your email list — a sale announcement to owned audiences costs nothing per impression and does not compete in an inflated auction, which is the whole argument in email marketing for Singapore businesses.
On the day — watch, do not tinker. Monitor stock, delivery status and site speed. Resist restructuring. Every structural change on the day resets something.
After — measure over fourteen days. A single-day revenue figure flatters every sale, because a large share of it is demand pulled forward from the following weeks and none of it is net of returns. Judge the event on a fourteen-day window against the fourteen days before, and on margin rather than revenue. ROAS versus ROI is exactly the distinction that decides whether a 40%-off day was a good idea.
What to do if you are not an electronics or furniture retailer
The data is blunt about this. If you sell groceries, everyday consumables or services, the November spike largely is not yours, and bidding into the same auction as the categories that do spike means paying their prices for your traffic.
Three sensible responses:
- Sit out the peak days and buy the shoulders. Auction pressure is highest on the sale dates themselves. The first week of November and the week after 11.11 are materially cheaper and your buyers are still shopping.
- Sell a different thing. A service business can run a gift-voucher or prepaid-package offer that fits the season without competing on product discounts. It also collects cash now against delivery later, which is a better trade than a margin cut.
- Use owned channels. Email, WhatsApp broadcast to consented lists and organic social carry no auction cost. Note that consent rules differ by channel — PDPA and marketing tracking sets out the boundaries, and the Do Not Call provisions cover telephone numbers rather than email.
Five ways Singapore sellers lose money on these dates
- Discounting into a margin that was never there. Model the discounted unit economics including payment fees and returns before committing, not after.
- A strike-through you cannot evidence. Inflating the reference price in the weeks before is the specific practice CCCS’s discount guidance addresses. Keep the price records or drop the comparison.
- Tripling the Meta budget the day before. A significant edit resets learning at the worst possible moment.
- Applying a seasonality adjustment across a whole month. Google’s own guidance caps the useful window at 1 to 7 days and warns against periods over 14.
- Judging the result on sale-day revenue. It is the one number guaranteed to look good, and it is mostly pulled-forward demand.
Where this leaves you
The Singapore mega-sale calendar is real, measurable and considerably less symmetrical than the marketing around it suggests. November is the peak. September is the surprise second place. October underperforms both. December is a shop-floor month wearing an online costume.
Plan against that shape rather than against four equally-weighted dates, keep the price records that make your discounts defensible, and stop editing your ad accounts three days before the day you care about. That is most of it.
If you want the campaigns built and run properly across the season, that is what our performance marketing services do, and the outcomes are documented in our client case studies.
Where to go next
- Black Friday for Singapore e-commerce — the cross-border half of the November peak, and the S$400 rule that bites when baskets get bigger.
- Chinese New Year marketing in Singapore — the next seasonal peak after 12.12, and why the 2027 window is shorter than the last one.
- Your 2027 marketing budget — how to set the annual number these sale dates are spent out of.
- The performance marketing guide for Singapore — the pillar this sits under
- Digital marketing for Singapore e-commerce — the year-round programme
- E-commerce conversion rate optimisation — where sale traffic is won or lost
- Reducing cart abandonment — acute during a high-traffic sale
- Retargeting and custom audiences — how to use the browse traffic a sale generates
- Setting a Google Ads budget — before you scale it for a sale
Frequently asked questions
Is 11.11 or 12.12 bigger in Singapore?
On the official data, November. The Department of Statistics puts the online share of retail sales excluding motor vehicles at 19.3% in November 2025 against 17.0% in December, 16.7% against 15.3% in 2024, and 17.0% against 14.5% in 2023 — November ahead in all three years. December total retail is large, but a greater share of it happens in physical stores. If you must choose one month for concentrated online spend, choose November.
Is 9.9 worth running, or should I save the budget for 11.11?
Worth running. The online share of Singapore retail steps up sharply from August to September in both 2024 and 2025 — 14.3% to 16.2%, and 15.2% to 17.6% — making September the second-strongest online month of the year. October then falls back below September in both years, which makes 10.10 the weakest of the four dates on this evidence. Treat 9.9 as a real event and 10.10 as optional.
Can I raise my prices before a sale and then discount them?
Not safely. CCCS’s Guidelines on Price Transparency require that a discount or a comparison against your usual or previous price represents a genuine price benefit with a valid basis, and CCCS encourages suppliers to keep records of past sales and prices to prove that past prices and discounts were genuine. A reference price that was inflated shortly before the sale, and at which you did not meaningfully sell, has no valid basis. If you cannot evidence the “was” price, use a straightforward “$X off” or a bundle instead.
Can I advertise “no GST” or “GST absorbed” during a sale?
Not if you are GST-registered. IRAS states that a GST-registered business intending to give a discount equal to the GST amount should not advertise that there is “no GST”, because GST is still included in the post-discount price and must still be accounted for. GST-registered businesses must also display GST-inclusive prices on all public price displays, including websites and advertisements, with a fine of up to $5,000 for non-compliance. Advertise the equivalent discount instead.
Should I increase my Meta budget on the day of the sale?
Not on the day. Meta lists budget changes among the significant edits that can send an ad set back into the learning phase, which it exits after roughly 50 results in the week after that edit. A large increase on 10 November destabilises delivery on 11 November and the ad set cannot accumulate 50 results before the event ends. Step the budget up over the preceding two weeks so the ad set is already out of learning at sale-day spend.
How should I measure whether a mega-sale campaign worked?
Over a fourteen-day window, on margin rather than revenue. Single-day sale revenue includes a large amount of demand pulled forward from the following weeks and is gross of returns, so it always looks good. Compare the fourteen days from the sale against the fourteen days before, net of returns and discount, and against the same window last year if you have it.
Sources: Singapore Department of Statistics, online retail sales proportion by industry, monthly series retrieved via data.gov.sg on 31 August 2026; CCCS Guidelines on Price Transparency, in effect 1 November 2020 under the Consumer Protection (Fair Trading) Act; IRAS guidance on displaying and quoting prices, and on GST for imported low-value goods; Meta Business Help Centre, “About the learning phase”; Google Ads Help on seasonality adjustments for Smart Bidding. Examples are illustrative. Figures and rules change — verify before relying on them.
