If you sell a low-basket, high-volume TikTok Shop seller in the Philippines running creator orders at average order values between ₱99 and ₱199, creator content is probably your largest source of orders and, quite possibly, your largest source of refunds at the same time. The Philippines restored a flat per-order processing fee on 1 September 2026, and two properties of it matter more than the amount. It is flat, so it does not scale with basket size. And it is not refunded when the order is returned, so a cancelled sale still carries the cost. On a ₱299 basket that is a rounding error. On a ₱99 basket it is roughly five per cent of the order before a single creator commission, transaction fee or service tax is counted, and low-basket is exactly where Filipino creator volume lives: sample-size sachets, accessory sets, phone trinkets, the impulse end of the catalogue that converts beautifully in a fifteen-second clip and contributes almost nothing once the fixed costs are subtracted.
Do the arithmetic on one ₱99 order before deciding anything. Fixed per-order fee: ₱5, and you keep it if the buyer returns the item. Transaction fee at the prevailing rate plus service tax on top: call it roughly eight per cent of the basket, so about ₱8. Creator commission at a normal affiliate rate for the category: ₱10 to ₱20. Fulfilment for a small parcel: ₱25 to ₱45 depending on the zone and whether the buyer is outside Metro Manila. Add those and you are between ₱48 and ₱78 against a ₱99 basket, before the product itself is paid for. If your landed cost is 45% of selling price, the order is already negative — and it is negative in a way that grows with volume, which is the part that makes it dangerous. Selling more of that SKU through creators does not fix the unit economics. It multiplies the loss.
Where the fixed cost actually bites
A flat fee is not a cost problem, it is a threshold problem, and the threshold is arithmetic rather than opinion. Divide the fixed cost by the share of the basket you are willing to give it and you get the basket size. At three per cent, ₱5 implies a basket of about ₱167. At two per cent, about ₱250. Every peso of basket below that line is a peso where the fee, the transaction charge, the service tax and the fulfilment cost are all competing for a shrinking amount of room. The part sellers get wrong is treating this as a pricing question. Raising a ₱99 item to ₱129 does not solve it, because the fee stays at ₱5, the transaction charge and service tax rise with the price, and the buyer who was buying at ₱99 was buying at ₱99 for a reason — the clip that converts at ₱99 usually stops converting well before ₱129. The lever is the basket, not the price.
One more property of the fee makes low basket worse than the arithmetic alone suggests: it survives the return. On a returned order you have paid the fee, paid the shipping, and in many cases paid a commission that is difficult to claw back, and you have no revenue to set against any of it. So the true cost of the low-basket segment is not the fixed cost on the orders that stick, it is the fixed cost on the orders that stick plus the whole of the cost on the ones that do not. Sellers who model only the surviving orders consistently understate it, and the understatement is largest exactly where creator traffic concentrates, because creator traffic returns more than shelf traffic does.
Step 1 — Find the basket size at which the fixed cost stops mattering
Take last thirty days of creator orders and put every SKU on a line with four numbers: average basket, the fixed cost as a share of that basket, contribution per order after commission and fulfilment, and refund rate. Sort by the second column. Everything above whatever threshold you set — three per cent is a reasonable starting point for a fixed cost that also survives returns — is the list. Then look at the third column rather than the second. Some SKUs will sit above the threshold and still contribute well, usually because their refund rate is unusually low or their landed cost is unusually good. Those stay. The ones that are above the threshold and thin on contribution are the ones that have to move.
Use creator orders only, not store totals. A blended average hides the whole problem, because the shelf orders in the mix are usually at a higher basket and drag the average up. If your dashboard will not separate them, pull the affiliate settlement file instead — it has the order-level rows you need, and it is also the file that shows commission already paid on orders that were later refunded, which is the second number this exercise needs.
The sheet is where most people stop, because joining basket size to per-order cost to refund rate by hand is slow and easy to get wrong. If you want the structure ready, register with DAMI and ask the team for the free low-basket contribution calculator: it computes the fixed-cost share per SKU, flags everything above the line, and shows contribution per order beside it.
Step 2 — Move low-basket SKUs into bundles rather than raising prices
Bundling is the only move that keeps the price point and lifts the basket at the same time, and it works because the fixed cost is per order, not per item. Three sachets at ₱99 becomes ₱249 against one ₱5 fee, and the fee share drops from roughly five per cent to two. Two rules make a bundle convert instead of confuse. It has to be one decision, not a menu — a creator cannot sell 'choose any three' in fifteen seconds, but can sell 'this is the three-week set'. And it has to be visibly cheaper than buying the pieces, because the shopper is comparing against the single unit they came for. Where a bundle would push the parcel into a higher shipping bracket or a different zone rate, check the new fulfilment number before you commit — a bundle that lifts the basket by ₱150 and adds ₱60 of shipping has only moved half the problem.
Give the creator one bundle to sell, not a catalogue. The conversion lift from a single clear offer is usually larger than the lift from choice, and it keeps the brief short enough that the price and the contents both make it into the video. Watch one number to judge whether it worked: average basket on creator orders, not units sold. A bundle that sells fewer units at a higher basket is doing exactly what it was designed to do, and judging it on unit volume will make you cancel a change that is working.
Step 3 — Take what cannot be bundled out of the creator programme
Some SKUs cannot be bundled. A phone charm is bought as one; a replacement part is bought because something broke; a single-use item has no natural second unit. For those, the honest answer is that creator traffic is the wrong channel at this basket size, not that the commission is set wrong. Moving them out of the creator programme is not a retreat — the shelf, search and shop tab still sell them, and they sell them without a commission line attached. Keep them available, keep them in bundles where there is a natural pair, and stop paying a creator to sell a basket that cannot carry the cost of being sold. Write the rule down in advance: any SKU whose creator basket sits below your threshold for two cycles and has no bundle partner goes back to the shelf.
The exception worth keeping is a SKU that loses money on its own order and makes money on the customer. If a ₱99 creator order reliably brings a repeat purchase at a higher basket within the quarter, the acquisition is worth more than the order, and cutting it on unit economics would be a mistake. You can only claim this if you can measure it, which means repeat rate by acquisition channel rather than a general repeat rate. If you cannot separate them, treat the SKU as a loss and move it — an unmeasured exception is just a reason to avoid the decision.
Step 4 — Reprice the creator side for the baskets that stay
For the baskets that stay, reprice the creator side rather than the product. Three changes do most of the work: commission on a tier rather than a flat rate, so the SKU that can only afford 8% is not carrying the rate set for a SKU that can afford 20%; a minimum basket condition in the campaign, so the creator is paid against bundles rather than single units; and settling on a cycle that lags the return window, so commission is not paid out on orders that are about to come back. None of these reduce what a creator earns on a good order. They stop you paying full rate on the orders that were never going to survive.
Tell the creators before you change it, and tell them in terms of what they get. A tier is easier to accept than a cut, because a creator who sells bundles earns more per order than one who sells singles, and the ones who were already selling bundles should be the ones earning more. Publish the tier in the campaign rather than negotiating it per creator, so the same rule is visible to everyone and nobody has to guess whether someone else got a better rate.
How DAMI reads contribution per creator order instead of per store
DAMI brings an 8-million-plus creator database, competitor and competitor-store creator search, automated outreach tasks, AI-assisted outreach, sample management, campaign management, store analytics, and multi-store team dashboards into one place. None of this is visible in a store-level report, because the store looks fine while one slice of it is losing money on every order. Three steps put the slice on screen.
Step 1 — Separate creator orders from everything else
Read affiliate orders as their own set rather than as part of the store total, with basket size and refund rate attached. Output: one view that shows the average basket on creator orders, which is almost always lower than the blended figure you have been quoting.
Step 2 — Put contribution per order next to basket size
For each SKU, show average basket, the fixed cost as a share of it, and contribution after commission and fulfilment in the same row. Output: the SKUs that are above your threshold and thin on contribution surface in one sort, without joining files by hand.
Step 3 — Watch the basket move after the change
Track average basket on creator orders weekly for the first month after you bundle or delist. Output: you can see whether the change lifted the basket or just removed volume, while there is still time to reverse it.
Frequently asked questions
Is the per-order fee refunded when the buyer returns the item?
No. That is the property that changes the maths, not the amount. A returned order leaves you with the fee, the shipping both ways, and in most cases a commission that is hard to recover, against no revenue. Any model of low-basket economics that only counts surviving orders will understate the cost, and the gap is largest in creator traffic because creator traffic returns more.
Should I just raise the price instead of bundling?
Usually not. The fee is flat, so raising the price does not reduce its share as much as bundling does, while transaction fee and service tax rise with the price. More importantly, the buyer converting at ₱99 is converting at that number for a reason, and conversion on a clip tends to fall well before the price reaches a basket that makes the fixed cost comfortable.
What if my category has no natural bundle?
Then the SKU does not belong in the creator programme at this basket size. That is a channel decision, not a failure. Shelf, search and the shop tab still sell single units, and they sell them without a commission line, which is often enough to make a thin SKU worth keeping in the catalogue but not worth paying a creator for.
Can I track all of this in one spreadsheet?
For one cycle, yes. It stops working when the question changes from 'what is my average basket' to 'which SKU is above the line this week, and did the bundle I launched last week move it'. That needs the basket, the fee share and the contribution refreshed on the same rows each week, which is where manual sheets drift — not because the maths is hard, but because nobody redoes the join every Monday.
Conclusion
Every cost line in this article moves without anyone deciding to move it, and every one of them is cheaper to fix before the next campaign than after it. DAMI brings an 8-million-plus creator database, competitor and competitor-store creator search, automated outreach tasks, AI-assisted outreach, sample management, campaign management, store analytics, and multi-store team dashboards into one place.
Fifteen minutes, one decision: pull last month's affiliate orders, sort by average basket ascending, and look at the bottom ten SKUs. For each, write down whether it has a natural bundle partner. The ones that do not, and that sit below your basket line, come out of the creator campaign today — not next month, because every day they stay in it is another day of paying a flat fee on an order that cannot carry it.
Start with DAMI and run the next cycle from numbers you can defend.
