Growing marketplace revenue does not establish that a brand is keeping more money. More discounts, a lower-margin product mix, returns, fulfilment costs or advertising can consume the additional sales. The first useful question is not whether the dashboard looks better. It is which products and decisions changed the economics.
This guide gives brand founders and ecommerce teams a practical review sequence. It complements marketplace management and commerce intelligence; it is not a universal margin benchmark or a forecast of client results. The calculations below are illustrative, not Digital Bees case-study figures.
1. Agree what profit means in this review
Use a consistent contribution definition before comparing periods. Start with net recognised sales on an agreed basis, then deduct product cost and the variable costs attributable to those sales. Depending on the selling model, these can include marketplace charges, fulfilment, payment costs, logistics, returns and brand-funded promotions. Deduct advertising once, separately, if the contribution figure is before advertising. Fixed overhead and agency costs may need a further layer.
Make the treatment of taxes, cancellations, refunds and discounts explicit. Do not deduct a promotion again if net sales already reflects it. Avoid comparing a marketplace settlement, gross order value and an advertising-attributed revenue number as if they were the same thing. Contribution is a management view with stated inclusions; it is not automatically accounting net profit.
2. See how stronger sales can leave less contribution
Consider two hypothetical equal-length periods in the same currency and on the same sales basis. The example deliberately keeps advertising as a share of total sales unchanged.
- Period A: net sales 100,000; contribution before advertising 35,000; advertising spend 20,000; contribution after advertising 15,000.
- Period B: net sales 150,000; contribution before advertising 37,500; advertising spend 30,000; contribution after advertising 7,500.
- Sales rose 50%, while contribution after advertising fell 50%. Advertising spend divided by total sales stayed at 20% in both periods.
- The pre-advertising contribution rate fell from 35% to 25%. That change, not the unchanged spend-to-sales ratio alone, explains why the business retained less contribution.
This is arithmetic, not evidence that advertising caused the margin decline. The next investigation should test product mix, costs, discounting and returns. Fixed costs, cash collection and working capital are outside this simplified example.
3. Read advertising metrics with the correct denominator
ACoS compares advertising spend with advertising-attributed sales. ROAS expresses advertising-attributed sales relative to spend. A total-advertising-cost-of-sales view, often called TACoS, compares advertising spend with total sales on a stated basis. Keep attribution windows and report dates visible. None of these ratios, alone, proves incremental demand or net profitability.
Amazon Ads explains that break-even ACoS relates to margin. A simplified break-even calculation is useful only when the sales and variable-cost bases are compatible and the included costs are understood. Do not apply a blended account margin blindly to every SKU or treat an attributed order as an order caused entirely by advertising. Branded demand, repeat customers and different campaign roles need context.
4. Reconcile at product level before changing budgets
Account totals can conceal a profitable product subsidising an unprofitable one. Build a view by SKU, platform and comparable period. Separate what is observed from the hypothesis explaining it. If margin or return inputs are missing, flag the gap rather than presenting a precise profitability conclusion.
- Sales and units: comparable dates, selling model and net-sales definition.
- Economics: product cost, marketplace and fulfilment charges, promotion treatment, returns and pre-ad contribution.
- Media: spend, attributed sales, campaign role and attribution limitations.
- Retail readiness: price, product-page quality, offer eligibility and actual availability.
- Action: the proposed intervention, accountable owner, review date and evidence needed to judge it.
5. Distinguish demand problems from buying barriers
A fall in sales can reflect fewer relevant visits, weaker conversion or an inability to buy. Check sessions, product-page information, offer conditions, stock and fulfilment availability before assuming the answer is more traffic. For quick commerce, availability can vary by location: a product listed nationally may not be purchasable in the demand zones being reviewed.
Treat unavailable locations as an operating constraint, not a calculated lost-sales figure unless a defensible method supports that estimate. Avoid claiming advertising waste from stock status alone when the actual campaign delivery and location coverage are unknown. Read more about the availability and visibility operating model.
6. Make the next decision small enough to evaluate
The first action might be repairing an incorrect listing, resolving a stock gap, investigating an unexpected fee or testing a narrower product allocation. A budget increase should have a commercial hypothesis, a spend limit and a review point. A budget cut also needs context: acquisition campaigns, launches and repeat-demand campaigns may have different objectives.
In Digital Bees’ operating approach, automation can help surface repetitive monitoring signals while commerce specialists reconcile evidence and prioritise action. That does not mean every cause can be identified automatically or that a particular improvement is guaranteed. Scope depends on authorised data, platform access and the brand’s responsibilities.
A useful weekly decision record
- Observation: what changed, for which products and over which dates?
- Evidence quality: which reports agree, and which costs or outcomes remain unknown?
- Decision: what will change, why, and who owns implementation?
- Guardrail: what spend, inventory or contribution condition would trigger a review?
- Outcome: what happened after enough comparable data accumulated? Record association without assuming causation.
Evidence and next steps
Our published India-market skincare case study describes marketplace work and reported growth metrics. It does not disclose contribution-margin results and must not be used as proof of a profitability uplift. A brand-specific review begins with its own data and operating conditions.
Explore retail media and marketplace advertising or commerce intelligence if sales, media and margin reports are pointing in different directions. Agree the available data and decision scope before commissioning work.
Sources and calculation note
Source checked 26 October 2026: Amazon Ads: advertising cost of sales. The contribution example and review worksheet are Digital Bees editorial frameworks. Illustrations exclude fixed costs and do not represent a client result, financial forecast or universal performance target.