AED 125.9K in Revenue. A 29.5% Account Margin. What Was Hiding Underneath?

A 29.5% account-level contribution margin looks reasonable. Not remarkable, but not alarming either. A team reviewing the headline might conclude: the account is in a healthy position. Consider scaling.

But account-level margin averages together every product. Strong margins from well-performing products raise the number. Weak or loss-making products get absorbed into it. The average tells you almost nothing useful about which products deserve more investment and which are quietly draining what the better products earn.

When we ran a 30-day SKU profitability diagnostic across this UAE account — covering Amazon and Noon across 39 product families — the same 29.5% account margin contained 4 loss-making families and 8 more below 15% margin. It also revealed one child SKU on Noon whose advertising spend exceeded its total revenue during the period.

The headline number was not wrong. It was just incomplete.

The key takeaway

Account-level margin is useful as a monitoring signal, not as an operating decision. Before scaling spend or inventory, look at which product families are actually contributing and which are consuming what others earn.

• 29.5% account marginmasked 4 loss-making product families and 8 more below 15%.
• Amazon and Noonshowed nearly identical account margins — but at product level, the same SKU could be profitable on one and loss-making on the other.
• One child SKUhad advertising spend that exceeded its total marketplace revenue during the period.

The account-level economics

Over the 30-day period, the account produced the following across Amazon UAE and Noon combined:

That 29.5% is a genuine average. It is not misleading on its own terms. The question is what it obscures.

What 29.5% was averaging over

The same account, broken down to product-family level, produced a very different picture.

The loss-making families were not large enough to push the account into an overall loss. They were small enough to disappear inside the average. That is precisely the problem with managing from a blended number — the products that most deserve attention are the ones most likely to be hidden.

Amazon and Noon showed almost the same account margin — and that was not useful either

One pattern in this account was particularly deceptive at first glance.

Two marketplaces. Contribution margins within a fraction of each other. At account level, there was almost no meaningful difference between them.

The first temptation when you see this is to conclude: Amazon and Noon are performing equivalently. Optimise both the same way. That conclusion breaks as soon as you look at individual product families. The same product that contributed 22.3% on Amazon contributed −22.4% on Noon. The account-level margins being similar was not evidence that the same products were working on both channels. It was evidence that the losses on one were being offset by gains on another — in a way that happened to cancel out to the same number.

What a single product family looked like underneath

One product family generated approximately AED 6,167 in combined revenue over the period. Its combined contribution margin was 10.4%. At the account level, that number would contribute to lowering the overall average — it looks like a below-average but survivable product.

Combined revenue: AED 6,167 · combined margin: 10.4%

The pre-advertising margin on both channels was nearly identical — approximately 32.7–32.8%. Advertising then consumed 10.5% of total revenue on Amazon and 55.1% on Noon. The product cost and fee structure were not the primary cause of the difference. Advertising was.

That is a more useful diagnostic than "this product has a 10.4% margin." It identifies exactly where to investigate and what kind of change might be available. The question becomes: why is Noon advertising consuming 55.1% of this product's total revenue, and what should change? That is a campaign and bid question, not a product question. The economics would look similar if advertising were brought into a range comparable to Amazon. The product itself is not broken on Noon — the advertising economics are.

One SKU had advertising spend greater than its total revenue

Within the loss-making Noon segment, one child SKU showed an extreme version of the same problem.

Noon child SKU · 30-day period

136.6% TACoS means advertising spend exceeded total marketplace revenue by approximately AED 237. Contribution margin for this SKU was approximately −106.9%. This did not appear in the account-level view. It appeared only at child-SKU level.

To be clear: this single SKU was not large enough to materially damage the account on its own. Its revenue was small. But it illustrates what can sit inside an account that looks healthy at the summary level. And the question it raises is straightforward: how many more SKUs like this exist, and what is the total capital being misallocated? The diagnostic exists to find them before the numbers grow larger.

The account also contained genuinely strong products

The point of a SKU-level diagnostic is not to find problems and ignore everything else. It is to separate the portfolio — so the business knows which products to protect, which to fix and which to stop pushing. Two product families in the same account showed what genuine performance looks like:

Both produced solid margins at meaningful revenue. Those products — and others like them — are the ones that deserve additional budget, inventory priority and advertising attention. But to make that decision confidently, you first need to know that you are not simultaneously scaling the products that are consuming what they earn.

The capital question is not simply: which product has the best ROAS? It is: which product leaves the most per dirham invested, across the full cost stack including product cost, marketplace fees, advertising and returns?

Four different operating decisions came out of the diagnostic

Once the product families were separated by contribution margin and marketplace, the next actions became clearer.

The diagnostic is a starting point, not a final answer

A SKU-level contribution analysis gives you the map. It does not automatically tell you what to do with every product. That decision requires understanding each product's stage in the marketplace lifecycle — launch, growth, maintenance, exit — strategic importance to the range, inventory position, pricing flexibility, advertising improvement potential, customer lifetime value where applicable, and marketplace-specific competitive dynamics.

But without the diagnostic, those conversations happen blind. The team may be investing in products it cannot see are loss-making. It may be under-investing in products it cannot see are the strongest earners. The 29.5% average does not help separate them.

When should a marketplace account run this kind of diagnostic?

This diagnostic is most useful in three situations.

Before a scaling decision. If the business is planning to increase advertising spend, expand inventory, add new products or enter a new marketplace, knowing which products are actually profitable at the current scale should precede any of those investments.

When the account is performing in line with expectations but the business is not growing profitably. A stable overall margin can mask a worsening composition — strong products holding the average up while the number of loss-making SKUs quietly grows. The diagnostic finds this before it becomes a larger problem.

Periodically as a standard operating review. Not every quarter needs a deep diagnostic. But an annual or semi-annual product-level review should be a standard part of running a multi-product, multi-marketplace account. Account-level reporting should be supplemented with at least a product-family view on a regular cadence.

What this case does not claim

• The account was unprofitable — account-level contribution was positive at 29.5%.
• The loss-making product families should automatically be removed — some may be launches, trials or strategic holds. The diagnostic identifies them; the decision requires context.
• Fixing Noon advertising alone would make the loss-making SKUs profitable — advertising is one input. Pricing, fees, product cost and returns all affect the result.
• The 30-day window is representative of a full year — seasonal, promotional or product-lifecycle effects could change the composition over a longer period.
• Contribution margin equals accounting profit — it excludes overhead, shared costs and other charges outside the analysis.

Frequently asked questions