Shopee, Lazada and TikTok Shop Fee Changes: Protecting 2027 Margins

If your marketplace margins are visibly thinner than they were two years ago and you cannot point to the reason, the cause is almost always cumulative rather than a single announcement. Category commission, transaction and payment fees, free-shipping programme contributions, campaign and voucher co-funding, and advertising cost have each moved independently, in separate seller-centre notices, and nobody rebuilt the per-SKU cost sheet after any of them. The fix before 2027 is not a negotiation with Shopee, Lazada or TikTok Shop, because an SME will not get one. It is an honest per-SKU, per-channel contribution calculation drawn from your settlement reports, a minimum margin floor you actually enforce when repricing, and a decision about which products you are willing to keep listing at the new rate.

What is actually changing in marketplace fees going into 2027?

Platform economics in this region have shifted from buying growth to defending profitability, and that shows up on the seller side as a thicker stack of deductions rather than one headline rate. In practice, a Singapore seller is now paying some combination of category commission, a payment or transaction fee on gross order value, a contribution toward the free-shipping or subsidised-delivery programme, co-funded platform vouchers during campaigns, return and reverse-logistics handling, and advertising that has become close to mandatory for visibility on TikTok Shop in particular.

Each of those is revised on its own schedule and published where most owners never look: the seller-centre announcement panel. Nobody emails you a profit-and-loss impact statement. The practical first step is unglamorous. Open each seller centre, download the current fee schedule for the categories you actually sell in, save it with the date in the filename, and keep it. Without a dated baseline you cannot tell a fee increase from a product mix shift, and you will spend 2027 arguing with your own numbers.

Why does a two-point commission rise hurt more than it looks?

Because fee increases land on contribution margin, not on revenue. Consider an item selling at S$40. Cost of goods is S$18. Platform fees at a combined 12 per cent take S$4.80. Your share of shipping subsidy is S$2. Packing, labour and last-mile handling add S$1.50. Advertising at 10 per cent of order value takes S$4. Contribution is S$9.70, or about 24 per cent.

Now add two percentage points of commission. That is S$0.80 on a S$40 order, which sounds trivial, but it removes a little over 8 per cent of what the SKU actually contributes. Add a one-point rise in payment fees and a modest increase in cost per click, and you are down 15 to 20 per cent of contribution on a product whose selling price has not changed at all. This is the arithmetic that quietly turns a profitable catalogue into a busy, high-turnover, break-even operation. Volume keeps rising, the warehouse keeps getting busier, and the bank balance stops moving.

How do you work out your true cost per channel?

Build one sheet with a row per SKU per channel, and resist the temptation to do it at category level. The columns that matter are selling price, cost of goods, commission, transaction fee, shipping subsidy exposure, voucher co-funding, allocated advertising spend, expected return rate multiplied by the cost of a return, and fulfilment labour in minutes multiplied by a realistic hourly rate including CPF.

Critically, populate the fee columns from your settlement or payout reports, not from the published fee schedule. The settlement report is the income statement of your marketplace business. It shows what was actually deducted, including the campaign fees and adjustments that never appear in the headline rate card. Reconcile one full month of payouts against orders for each channel. If the reconciliation is more than a couple of per cent out, you have found something worth more than this entire exercise.

Run the same calculation for your own website and for direct or WhatsApp orders. Most SMEs discover their own channel carries a far better contribution margin than they assumed, because they were comparing marketplace revenue against website revenue instead of comparing what each channel leaves behind.

Which SKUs should you stop listing?

Sort the sheet by contribution per unit and you will get three groups. The first is negatively contributing lines, which must be repriced, repackaged into bundles that clear the floor, or delisted. The second is thin but genuinely strategic lines that pull traffic or clear ageing stock, which you keep deliberately, with capped advertising and a review date. The third is healthy lines, which deserve stock priority, inventory cover and defended ranking.

The group that needs scrutiny is the second. It brings traffic is the most expensive sentence in Singapore ecommerce when it is said without a number attached. If a loss-leading SKU is justified by the follow-on purchases it generates, you should be able to show the repeat-purchase rate of customers acquired through it. If you cannot, it is not a strategy, it is a habit.

What can you realistically fix in the rest of 2026?

There are under ninety days left in the year and the next fee cycle, Chinese New Year demand and January cost step-ups all arrive together. A workable sequence: download twelve months of settlement reports per channel this week; rebuild per-SKU contribution for your top eighty per cent of revenue; set a written minimum contribution percentage and apply it to every price change from now on; review supplier cost and minimum order quantities before CNY production slots close; and reduce single-channel dependence by giving customers one reason to order directly, whether that is a WhatsApp reorder flow, a loyalty price, or proper invoicing for your B2B buyers.

Then assign one named person to read the seller-centre announcements monthly and record changes in a shared log. Fee changes are published, missed, and absorbed. That sequence is a process failure, not a platform conspiracy, and it is cheap to fix.

Does this mean leaving the marketplaces?

For most sellers, no. Marketplaces supply demand you cannot buy more cheaply anywhere else, and walking away to protect pride usually costs more than the fees did. The mistake is treating a marketplace as the whole business and letting fee drift set your margin by default. Decide what proportion of revenue you are comfortable having inside platforms you do not control, treat anything above that as concentration risk alongside your customer concentration risk, and price every listing against a floor you chose rather than against a competitor who may not know their own costs. That is the difference between selling on a marketplace and being owned by one.

Frequently asked questions

How often do marketplace seller fees change in Singapore?

There is no fixed schedule, which is precisely the problem. Category commissions, payment fees and shipping programme contributions are each revised independently and announced through the seller centre, often with a few weeks of notice. Treat a monthly ten-minute check of announcements across every platform you sell on as a standing operational task, the same way you check bank reconciliations.

Should I just raise my prices to cover higher fees?

Only after you know your contribution by SKU. Blanket increases push your best-selling, price-sensitive lines out of the buy box while leaving genuinely underpriced products untouched. The better approach is selective: raise prices where you have differentiation or bundle value, hold on lines where ranking is doing real work, and delist what cannot clear your floor at any plausible price.

Can I track contribution margin properly without buying an ERP?

Yes, for a catalogue in the hundreds of SKUs, a disciplined spreadsheet rebuilt from settlement reports each month is sufficient and far better than an unused system. The case for consolidating into proper inventory and costing software appears when stock sits across several channels and warehouses, when manual reconciliation starts taking days, or when you cannot answer what a product costs you without asking two people. At that point the tooling question is worth a deliberate decision rather than another spreadsheet tab.