Key Takeaway
A marketplace account review should begin with the business question, not with a generic list of red metrics. Identify the signal that triggered concern, add the relevant product, advertising, inventory, returns or financial context, establish what the evidence supports, and then prioritise the issue by commercial impact, urgency, confidence and actionability.
Marketplace accounts rarely have just one problem. When several signals appear together, the natural response is to open more reports and react to more metrics. That usually produces more observations — and no clearer priority.
A useful account review moves in the opposite direction.
Different questions require different evidence
Before opening Amazon Seller Central, Noon reports, advertising dashboards or inventory sheets, ask: what are we actually trying to understand? The starting question determines which of these lenses becomes relevant. Not every review needs all six.
Lens 01 — Product Economics
Which products are actually worth more attention?
Revenue is usually the easiest number to see. It is not always the most useful number to prioritise. In one real UAE marketplace account, the number-one Amazon SKU by revenue generated approximately AED 12,276 in revenue and AED 2,800 in contribution at a 22.8% contribution margin.
Another SKU generated lower revenue — approximately AED 10,113 — but left approximately AED 3,359 in contribution at a 33.2% contribution margin. The best-selling product was not the highest-contributing product. That changes the question.
Instead of asking which SKU sold the most, ask which SKU deserves the next dirham of advertising, inventory or management attention.
In the top-selling example above, advertising represented only 7.4% of revenue, while the recorded marketplace-fee line represented approximately 46.7%. That meant "reduce PPC" was not automatically the first investigation. The better question was: what is creating the fee burden, and is any part of it something we can influence?
A useful product-economics view may include revenue, units sold, product cost, recorded marketplace fees, advertising and contribution. The purpose is not to create another profitability score — it is to understand what is actually consuming the economics of the SKU. See Blog 01 for a deeper look at this diagnostic.
Lens 02 — Advertising
Is paid media actually the problem?
Advertising is one of the easiest areas to overreact to. When contribution looks weak, the instinct is often to cut PPC. When ROAS looks strong, the instinct is to scale PPC. Both can be wrong.
One Amazon SKU in the same account showed 4.03x ROAS but only 1.7% contribution margin. The advertising metric looked respectable. The economics did not. ROAS tells you how much ad-attributed revenue was generated for the advertising spend. It does not tell you what happened after product cost, marketplace fees and other included costs were considered. See Blog 03 for the full analysis.
The same principle becomes even more important across marketplaces. Another product family generated approximately AED 13,453 across Amazon and Noon over 90 days at a combined contribution margin of only 8.0%. But split by marketplace: Amazon at approximately 18.2% contribution margin; Noon at approximately −19.7% contribution margin. The product itself had not suddenly become a different product. The marketplace economics had changed. See Blog 02 for the channel comparison.
When the Noon campaign was decomposed, some targets were still producing useful orders while others were spending without converting. The operating question became: what should we reduce, what should we protect, and what needs more evidence? That is a better account-review question than simply asking whether PPC is good or bad.
Lens 03 — Inventory
Do we need to buy more, or move what we already have?
Stock alerts can create the wrong action when viewed in isolation. In one inventory snapshot from the same operating environment: 38 SKUs were below reorder point. At the same time: 33 SKUs were overstocked. There were 1,510 units ready to ship and only 17 suggested reorders requiring approximately AED 20,096 in purchasing capital.
The inventory question that matters
Before committing more working capital, the review should separate two questions: Is the marketplace running low? and Does the business actually lack stock? Those are not the same problem. A low-stock alert tells you where to look — it does not tell you to buy.
One SKU had approximately 564 units across the network — including 400 units in the local warehouse. Amazon was consuming stock much faster than Noon. The recommended supplier reorder was 0 units. The more useful action was to move existing inventory: 100 units to Amazon and 50 units to Noon.
Before committing capital, the review should understand: current marketplace stock, warehouse stock, sales velocity, inbound inventory, open purchase orders, supplier lead time and whether existing inventory can be redistributed. See Blog 05 for the full inventory diagnostic.
Lens 04 — Returns
Are customers keeping what we are paying to sell them?
Returns deserve more attention than simply adding a percentage to a weekly dashboard. Across one 90-day account view: 413 units were returned. The blended return rate was 4.5%. Recorded return cost was AED 10,616.76. Those figures tell us returns are commercially relevant. They still do not tell us what to fix.
The diagnosis became more useful when the Amazon returns were broken down by reason. The leading recorded reasons included: Unknown, Quality Unacceptable, Unwanted Item, Not Compatible and Defective. One child SKU showed 60 units sold, 8 returns and a 13.3% return rate. Its recorded reasons included quality and compatibility signals, while one customer specifically complained that the product was leaking.
A good review should also be willing to conclude: we do not yet have enough evidence. The Noon return data in the same environment did not provide equivalent reason-level and customer-comment depth. It would therefore be wrong to use an Amazon customer complaint to explain Noon behaviour. See Blog 06 for the returns investigation framework.
Lens 05 — Amazon Reconciliation
Can the Amazon payout actually be explained?
Marketplace sales and bank receipts are different numbers. That sounds obvious. But when the gap becomes large, it can create immediate concern. In one Amazon settlement period: gross sales were approximately AED 43.8K. The actual payout was approximately AED 15.3K. At first glance, more than AED 28K appeared to sit between the two.
But after recorded fulfilment fees, referral fees, advertising, refunds, promotions, storage charges, other fees and positive adjustments were mapped, the remaining attribution difference was closer to AED 124. That changes the problem completely.
- Narrow a broad financial concern into a specific investigation
- Map each fee and adjustment category to its source
- Separate sales from payout from profit
- Isolate the true unexplained residual
- Identify whether that residual warrants further follow-up
- Residual variance is not automatically money Amazon owes
- A reimbursement record is not a confirmed claim
- A reconciliation gap is not a profitability conclusion
- Settlement timing can explain apparent gaps
- Reimbursement and reconciliation are separate questions
See Blog 04 for the full reconciliation walkthrough from AED 43.8K gross to AED 15.3K payout to AED 124 residual.
Lens 06 — Pricing, Listings & Conversion
Could the offer or product presentation explain the result?
Not every problem starts in advertising, inventory or cost. Sometimes the commercial metric changes because the product offer itself deserves investigation. Traffic can rise while conversion falls. A SKU can receive impressions but fail to convert. Customers may misunderstand compatibility or product use. A promotion may change sales without necessarily improving contribution.
So when a performance metric changes, the review may also need to ask: Has the price changed? Is the product competitive in its segment? Has a promotion altered the economics? Is the listing communicating the product clearly? Are important search terms missing? Is stock availability affecting conversion?
The important point is not that every account review should always conduct a complete pricing and listing audit. It is that these factors should enter the investigation when the evidence suggests they may help explain the original question.
Prioritisation
The biggest red number is not always the first problem to fix
A practical way to think about priority is through four questions. These prevent the review from becoming "fix everything that is red" — and turn it into "fix the things that matter, that we understand well enough, and that we can actually act on."
01
Commercial Impact
How much revenue, contribution, cash or customer experience is exposed? A problem affecting a major product or large amount of contribution deserves different attention from a small anomaly.
02
Urgency
What happens if nothing changes this week? A product approaching stockout may be more urgent than a small reconciliation variance that is already isolated and understood.
03
Confidence
Do we actually understand what is happening? A clear PPC waste pattern can justify action sooner than a return problem whose cause is still unknown.
04
Actionability
Is there something sensible the team can do now? Some issues can be acted on immediately. Others need more evidence first before any change is justified.
The important point: do not turn this into a numerical scoring formula. These are judgment dimensions — not a validated algorithm. The combination of all four is what matters, not a weighted average of scores.
A good review should end with a verb
The result of a review does not always have to be "optimise". One of the clearest signs of a useful account review is that it ends with a specific action rather than a longer list of observations.
Protect
What is already working and should not be disrupted?
Increase
What deserves more investment — advertising, stock or attention?
Reduce
Where is spend or stock exceeding what the economics justify?
Move
Can inventory or resources be reallocated rather than replaced?
Fix
Is there a clear underlying problem that deserves direct action?
Investigate
Is there a signal but insufficient causal evidence to act yet?
Hold
Would acting now create more risk or cost than waiting for better evidence?
A good review should not leave you with more metrics. It should leave you with fewer, clearer decisions.
Review Sequence
A practical marketplace review sequence
A useful account review can follow one simple sequence — ending with a priority, not simply a list of metrics.
What is the business question?
Be specific about what is concerning the team. "Sales are growing but the business feels less profitable" is more useful than "review everything."
Which metric or change triggered the question?
Revenue, contribution, ROAS, conversion, stock, payout, returns or another signal — identify the specific number that created the concern.
What context could change the interpretation?
Marketplace, SKU, objective, inventory, costs, customer behaviour, pricing or time period — what else should be in the frame before drawing a conclusion?
What does the evidence actually support?
Separate facts from assumptions. "The return rate is 13.3%" is a fact. "The product has a quality defect" may be an investigation route, not yet a conclusion.
What is the likely commercial exposure?
Understand whether the issue materially affects revenue, contribution, cash or customer experience before deciding how much time and resource to invest in solving it.
How urgent is it?
Some problems deteriorate quickly — a product approaching stockout, a return rate still being investigated during active advertising. Others can be monitored.
What action is available?
Reduce, protect, redistribute, reorder, investigate, fix, hold or gather more evidence — identify what is actually possible before committing.
What should happen first?
End the review with a priority, not simply a list of metrics. The question changes from "what is wrong?" to "what deserves attention first, and what should we do about it?"
Preparation
What should a marketplace team prepare before an account review?
You do not need to diagnose the account before asking for help. But the investigation becomes more useful when the available information matches the question.
| If the question is about… |
Information that may become relevant |
| Profitability |
Product costs, recorded marketplace fees, revenue and contribution by SKU |
| Advertising |
Campaign and target-level data, ROAS, TACoS, ad spend by SKU and marketplace |
| Inventory |
Current marketplace stock, warehouse stock, sales velocity, inbound and open POs, supplier lead times |
| Returns |
Return rate by SKU, reason-level or customer-feedback evidence, disposition data |
| Reconciliation |
Amazon settlement records, fee category mapping, payout history |
| Unclear / general concern |
Explain what changed, what looks unusual, and what the business is trying to achieve — the investigation can then determine what evidence is relevant |
Review access is not permission to make changes
A business may be comfortable allowing an account to be reviewed without authorising another party to change campaigns, pricing, stock allocation or marketplace settings. Those are separate decisions. Saddl first confirms what information or access is necessary for the question being investigated. Any permission to execute changes should be agreed separately.
This is where Saddl becomes useful
A marketplace business does not usually suffer from a lack of data. The harder problem is that the data sits in different places and answers different questions. Each view is useful — the value comes from connecting them around one decision.
Saddl · Investigation & Decision Model
01
Marketplace Signals
Sales, advertising, stock, returns, economics and relevant account records — each view answers a different question.
02
Saddl OS
Brings supported operating signals into connected views where the relevant data and capability are available.
03
Operator Context
Product role, objectives, market conditions, account history and commercial priorities — the business context only the operator knows.
04
Investigation
Question the likely causes and determine what the evidence actually supports — not what the dashboard implies.
05
Priority
Decide what deserves attention first — by commercial impact, urgency, confidence and actionability, not by which metric is reddest.
06
Execution
Agree actions, responsibilities and permissions separately. The review informs the decision. Execution is a separate step.
The objective is not more dashboards. It is fewer unresolved questions.
What should a good account review leave you with?
Not fifty issues. Not a score out of 100. Not a generic recommendation to optimise advertising, improve listings and manage inventory better. A useful review should leave the team clearer about questions such as:
What genuinely needs new inventory?
What deserves attention first?
That is the difference between reviewing an account and simply reading one.
Frequently asked questions
What is a marketplace account review?
A marketplace account review investigates a business question using the relevant account information and operating context. The goal is to understand what deserves attention and what a sensible next action may be, rather than simply producing a list of red metrics.
Does every marketplace account review cover advertising, inventory, returns and profitability?
No. The areas reviewed should depend on the question, account and information available. A profitability issue may require different evidence from an inventory or reconciliation question. Saddl's Account Review is deliberately shaped by the account and the question being investigated, rather than following a fixed universal checklist.
Do I need to know what the problem is before requesting a review?
No. It is enough to explain what is happening, which numbers are difficult to interpret and what the business is trying to achieve. The investigation can then determine what evidence is relevant and which areas deserve attention.
Does an account review mean changes will be made to my marketplace account?
No. Review access and permission to execute changes are separate. Saddl first confirms what information or access is necessary for the question being investigated. Any permission to execute changes should be agreed separately with the appropriate scope and permissions.
What is the most important metric to check in a marketplace account?
There is no single metric that answers every marketplace decision. Revenue, contribution, ROAS, TACoS, conversion, inventory, returns and settlement data all answer different questions. The useful metric depends on what the business is trying to understand — which is why the review should start with the business question, not with the dashboard.
How do I decide which marketplace problem to fix first?
Prioritise the issues with the strongest combination of commercial impact, urgency, confidence in the diagnosis and ability to act. A dramatic metric is not automatically the highest-priority problem. A clear, lower-impact problem that can be fixed today may deserve priority over a larger but poorly-understood issue where more evidence is needed first.
Deeper evidence behind each review lens
Each example in this guide comes from a dedicated evidence-led resource. Read the ones relevant to the question your account is raising.
A good marketplace review should leave you with fewer, clearer decisions
Tell us what is happening in your marketplace account and what your team is trying to understand. Saddl can investigate the relevant signals in context and help identify what deserves attention next.
Requesting a review does not authorise account changes or automatically start an ongoing engagement. Scope, information requirements and any later execution are agreed separately.
Request an Account Review
Stop Guessing. Start Executing.
Guide Note
The account situations used in this guide are anonymised examples drawn from real UAE marketplace operating data used across Saddl's resource analysis.
Individual examples use different reporting periods and datasets and should not be interpreted as one single account snapshot or as universal marketplace benchmarks.
Contribution refers to the operating contribution calculation used in the relevant analysis and should not be treated as statutory or accounting net profit.
Recommendations and diagnostic findings described in this guide do not imply realised financial outcomes unless an outcome is explicitly documented. Marketplace data depth and capability can differ by platform and account. Amazon reconciliation applies to Amazon settlement specifically. An Account Review does not universally promise every review lens shown in this guide — the account and question determine which areas become relevant.