Case Study · SKU Profitability

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 30-day SKU-level diagnostic across 39 product families found 4 loss-making families and 8 more below 15% hiding inside that same average.

UAE marketplace account · 30-day diagnostic · Amazon UAE + Noon · 39 product families

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.

AED 125.9K
Total revenue · 30 days
39
Product families
29.5%
Account margin
4
Loss-making families

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:

LineAmount% of revenue
RevenueAED 125,899
Ad spendAED 11,8469.4%
Product costAED 31,16524.8%
Marketplace feesAED 45,78236.4%
ContributionAED 37,105
Contribution margin29.5%account average

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.

4
Product families running at a loss — negative contribution after costs
8
Product families below 15% contribution margin
1
Noon child SKU where ad spend exceeded total revenue

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.

Amazon UAE

29.46%

Account contribution margin

Noon

29.51%

Account contribution margin

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.

Amazon UAE

+22.3%

10.5%
TACoS
~32.8%
Pre-ad margin

Noon

−22.4%

55.1%
TACoS
~32.7%
Pre-ad margin

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.

AED 647
Revenue
AED 884
Ad spend
136.6%
TACoS

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:

Product family A

41.3%

Revenue: AED 8,114 · 30-day period

Product family B

38.4%

Revenue: AED 11,600 · 30-day period

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.

01
Fix advertising on loss-making Noon SKUsWhere pre-advertising margins are similar to Amazon but TACoS is far higher, the campaign economics need attention before anything else is changed. The product itself may not need repricing or removal.
02
Scale the strong-margin familiesProducts at 38–41% contribution margin with meaningful revenue should receive advertising and inventory attention ahead of weaker ones. Blended ROAS targets should not be used to constrain these products.
03
Review the below-15% familiesEight families below 15% margin deserve individual review — not automatic removal. Some may be viable at higher price or lower advertising intensity. Some may be early-stage products. The diagnostic identifies them; the business decision depends on context.
04
Set meaningful margin floorsRather than managing against an account-level average, set contribution-margin targets at product-family level. Use those to decide which products should receive more budget and which should be reviewed.

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.

01
Account level29.5% contribution margin. Looks healthy. No obvious action required.
02
Marketplace levelAmazon 29.46% — Noon 29.51%. Almost identical. Still no obvious action.
03
Product family level4 loss-making families, 8 below 15%. Same product: Amazon +22.3%, Noon −22.4%.
04
Child SKU levelOne Noon child SKU: TACoS 136.6%, contribution −106.9%. Advertising spend exceeds revenue.
05
Operating decisionFix Noon advertising on specific SKUs. Scale strong-margin families. Review below-15% products individually. Set margin floors.

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

Why is account-level margin not enough to make decisions?

Account-level margin is an average. Strong products raise it, weak ones are absorbed into it. In this case, a 29.5% account margin contained 4 loss-making product families and 8 more below 15%. Those products would not be visible without a product-level breakdown.

What is TACoS and why does it matter?

TACoS — Total Advertising Cost of Sales — is advertising spend divided by total product revenue. It shows how much of a product's revenue is being consumed by advertising. In this account, one Noon child SKU had a TACoS of 136.6%, meaning its advertising spend exceeded its total marketplace revenue during the period.

Can the same product be profitable on Amazon but loss-making on Noon?

Yes. In this account, one product family generated a +22.3% contribution margin on Amazon and a −22.4% margin on Noon during the same 30 days. The pre-advertising economics were almost identical on both channels — the difference was that Noon advertising was consuming 55.1% of the product's total revenue versus 10.5% on Amazon.

Should loss-making SKUs be removed?

Not automatically. A loss-making result during a 30-day period could reflect a product launch, a promotional period, a campaign that can be optimised, or a product that genuinely cannot be sold profitably at the current price and cost structure. The diagnostic identifies which products warrant review. The decision requires understanding each product's context.

How often should a SKU profitability diagnostic be run?

That depends on the account size and rate of change. A quarterly review is common for active accounts. An annual or semi-annual deep diagnostic should be standard for multi-product accounts operating across multiple marketplaces. The most important time is before a major scaling or investment decision.

What does contribution margin include in this context?

In this analysis, contribution refers to revenue minus product cost, recorded marketplace fees and advertising spend. It is an operating decision metric. It does not include overhead, logistics outside the recorded fee structure, shared costs or other charges, and should not be interpreted as accounting profit.

The account average cannot tell you which products to scale

A 29.5% account margin is useful context. It is not operating information. To decide which products deserve more capital — and which are quietly consuming what the better products earn — you need the breakdown: by product family, by marketplace, by cost line. That is what a SKU-level profitability diagnostic produces, and what Saddl builds into standard account management so operators are not making scaling decisions from a number that averages over the products they most need to understand.

Request an account review →
Data note

This case study uses anonymised operating data from a real UAE marketplace account over a 30-day period across Amazon UAE and Noon, covering 39 product families.

Contribution in this analysis refers to: Revenue − product cost − recorded marketplace fees − advertising spend. It is an operating decision metric and should not be interpreted as accounting or statutory net profit. The SKU-level data provides aggregated recorded fee figures. Individual fee components are not separately identified in the available data, so no specific fee attribution has been made. The pre-advertising margin figures used in the product-family example are approximations based on the available cost data and should be treated as indicative comparisons rather than exact calculations.