Debates about whether marketing should be 15% or 22% of revenue often become contests between cost control and growth ambition. Neither percentage answers the real question. Leadership needs to know what the additional spend will buy, how much incremental contribution it will generate, and when the cash returns.
What remains from a VND 500,000 order?
| Order layer | VND per order |
|---|---|
| Revenue after 15% promotion | 500,000 |
| Platform fixed fee · 14% | (70,000) |
| Service fee · rounded to 0.3% | (1,500) |
| Transaction processing · 6% | (30,000) |
| Settlement after platform fees | 398,500 |
| Assumed COGS · 35% | (175,000) |
| Contribution before Ads and Affiliate | 223,500 · 44.7% |
| Assumed Ads · 8–10% | (40,000–50,000) |
| After Ads, before Affiliate and omitted costs | 173,500–183,500 · 34.7–36.7% |
All percentages use post-promotion product revenue, on an assumed consistent tax basis. The 15% promotion is already in the selling price: do not deduct it again. Net shipping cost is zero in the supplied settlement because a VND 56,800 charge was fully offset by an equal subsidy. That does not establish free shipping for every order.
Evidence: supplied settlement and modelling assumptions
| Settlement line | VND |
|---|---|
| Product amount after promotion | 938,818 |
| Shipping charge | (56,800) |
| Shipping subsidy | 56,800 |
| Fixed fee | (131,435) |
| Service fee | (3,000) |
| Transaction processing | (56,329) |
| Total fee deductions | (190,764) |
| Settlement receipt | 748,054 |
The settlement is user-supplied evidence, not independently audited. The VND 500,000 model scales and rounds its observed fees; 35% COGS and 8–10% Ads are assumptions. Affiliate is now assumed at 10% of net revenue for the illustrative orders; this is not a rate verified from the bill. The VND 3,000 service fee may be a per-order charge; its basis must be confirmed before using the scaled VND 1,500 for decisions. These are not a current or universal platform fee schedule.
| After Affiliate: the updated scenario | VND per order |
|---|---|
| Before Ads and Affiliate · 44.7% | 223,500 |
| Assumed Affiliate · 10% | (50,000) |
| Before Ads, after Affiliate · 34.7% | 173,500 |
| Ads · 8–10% | (40,000–50,000) |
| After Ads and Affiliate · 24.7–26.7% | 123,500–133,500 |
This scenario assumes the order incurs both Ads and Affiliate. Where only some orders incur Affiliate, use the actual commission-bearing revenue and the resulting effective rate, rather than charging 10% to every order.
01A Marketing Percentage Does Not Tell the Whole Story
At the same 20% of revenue ratio, one business may be acquiring genuinely new customers while another is subsidising orders that would have occurred anyway. High-margin and low-margin assortments also support very different affordable marketing levels.
Agree the scope of what is included: media, affiliates, agencies, content, samples, promotions and any other commercial investment. If “Ads & Promotion” is tracked narrowly, the remaining marketing costs must still be monitored. A partial metric should never be presented as total customer acquisition cost.
02Start with the Contribution Available per Order
The case retains VND 223,500 before Ads and Affiliate. With Affiliate assumed at 10% (VND 50,000), VND 173,500 remains before Ads. If leadership wants to retain VND 100,000 per order (20% of revenue), the provisional Ads ceiling is VND 73,500 (14.7%), before omitted costs. Equivalent ROAS is approximately 6.80× on the same net-revenue basis.
Deduct omitted variable costs from that ceiling. The 20% retention target and 10% Affiliate rate are assumptions, not industry benchmarks. The ceiling is a financial boundary, not an optimal budget or an automatic stop rule.
03Average ROAS Can Conceal Marginal Performance
Suppose an extra VND 10m of Ads causes VND 50m of incremental net revenue. At the illustrative 44.7% margin before Ads and Affiliate, contribution is VND 22.35m. Subtract incremental Affiliate at 10% (VND 5m) and extra Ads of VND 10m: VND 7.35m remains before omitted costs.
If the same spend causes only VND 20m of incremental net revenue, contribution after Affiliate and Ads is 20m × (44.7% − 10%) − 10m = −VND 3.06m, before omitted costs. Existing campaigns can remain profitable while the next tranche loses money. These are scenarios, not measured outcomes.
Use the economics of the incremental revenue. AOV, SKU mix, fees and returns may differ from the sample; 100 additional orders at VND 400,000 AOV yield VND 40m, not VND 50m. Recalculate the margin when these drivers change.
04Repeat Purchases Can Justify an Investment — With Discipline
A loss on the first order can be recovered through subsequent purchases. Measure customers acquired in the same period and track cumulative contribution after servicing and retention costs, not revenue as a proxy for profit LTV.
Separate genuinely new customers from existing ones. Where purchase history is short, label assumptions clearly as assumptions. Three weeks of data cannot validate a six-month repeat-purchase expectation. When the payback period lengthens, the cash required to fund the investment rises and must be planned explicitly.
05Separate Budgets by Purpose
| Purpose | Primary measure | Decision logic |
|---|---|---|
| Current-order generation | Short-term contribution after ads | Scale only if incremental contribution is positive |
| New-customer acquisition | CAC vs contribution LTV and payback period | Requires clear recovery thesis and cash plan |
| Brand / long-term demand | Reach, consideration, learning milestones | Not judged solely on short-term ROAS, but still needs hypothesis, ceiling and review points |
Brand investment should not be forced into a pure ROAS frame, yet it still requires a target audience, a testable hypothesis, a spending ceiling and defined learning milestones. Release funds for defined experiments first; scale only the approaches that produce stronger evidence. Weak results should trigger changes to creative, pricing or product before more money is committed.
06Give Leadership a Testable Choice
A proper marketing proposal should specify:
- Incremental spend requested
- Expected incremental contribution (not just revenue or ROAS)
- Cash timing and peak funding requirement
- Key assumptions and downside scenarios
- Clear stopping or review conditions
A 90-day review can assess conversion and early unit economics. Full validation of repeat-purchase assumptions depends on the product’s natural repurchase cycle. Finance and Marketing should jointly own the assumptions and the outcome review. The discussion then shifts from “15% versus 22%” to an accountable investment decision with defined success and failure criteria.
Key takeaway
The marketing percentage should follow the economics and the strategy. It cannot replace a clear investment thesis, evidence of incrementality, and disciplined conditions for releasing further funds.
The settlement was supplied for this editorial case; identifying details are excluded. Modelled COGS, Ads, Affiliate scenarios and thresholds are assumptions. This is an incomplete contribution model until omitted costs are added.
Back to E-commerce ↗