This is the report you get.
Every service in this category offers a free audit, and every one of them shows you the deliverable after you hand over account access. We could not find one that publishes the report first. So here is ours, end to end — the exact structure a client receives, filled with invented data on a seller who does not exist.
Read this before the numbers
Two things are true about this page at once, and it matters that you hold both.
The structure is real. Every column, every category, every refusal line below is what actually lands in a client’s findings report. Nothing has been simplified for the website, and nothing embarrassing has been taken out — the refusals and the undocumented SKUs are on the page because they are on the report.
The data is invented. The seller is fictional: a mid-size home goods brand, roughly 340 active ASINs, about $4M a year through FBA. The SKU codes are made up. The dollar amounts are made up. Every type of finding on this page is a type we genuinely find; not one of the figures came from anyone’s account.
Sample data, stated plainly. No client account, no real seller, no real ASIN, no real claim outcome appears anywhere on this page. Do not read the total as a forecast for your catalog — a 340-SKU home goods brand and a 40-SKU electronics brand do not have the same error profile, and neither of them has yours.
The reason this page exists is narrow and slightly petty: “audit your auditor” is the entire industry’s pitch, and the audit is a sales call. Asking you to trust the artifact without ever showing you the artifact is a strange thing to build a business on. This is the artifact.
What the audit covered
The first page of the report is scope, because a recovery number without a denominator is a marketing number.
340 ASINs
Mid-size home goods brand, ~$4M annual FBA revenue. Fictional, for this sample.
12 months
Full transaction, inventory, returns, removal and storage ledgers. Only findings still inside a claim window are listed.
How to read the columns
- SKU and category. One line per finding, not one line per SKU. A SKU with a dimension error and a lost-unit adjustment appears twice, because they are two claims with two deadlines.
- Evidence. What we would attach to the case. If this column is thin, the finding does not get filed — see what we refuse to file.
- Recoverable. What is claimable with that evidence attached. Not what will arrive. Amazon decides that part.
- Days left. Against the deadline for that finding type — 60 days for lost and damaged, 60–120 days after the refund for returns, 15–75 days for removals, nine months from creation for inbound shipments, and the separate 90-day fee-dispute deadline for the fee rows. It is the only column that gets worse while you think about it.
$6,517 recoverable, across 14 findings
Sorted by size, so you can see immediately where the money actually is. On this sample account the fee rows are 49% of it, which is the shape we see most often and the opposite of how this industry sells itself.
Sample figures on a fictional account. Not a quote, not a projection, and not what your catalog contains.
The number that should actually worry you
Total recoverable is the number every service leads with. It is the wrong one to look at first, because a claim outside its window is worth exactly nothing regardless of how large it is. So the second block on the summary page is the clock.
Nineteen percent of this account’s recoverable balance is on a two-week fuse. That is the argument against a quarterly reconciliation in one line, and it is why the report leads with expiry rather than with the headline.
Recoverable is not recovered, and we will not pretend otherwise. This column is what is claimable with evidence attached. Amazon approves or denies it, and it denies plenty. On our own account we published a 19% win rate on storage claims — two of the fourteen findings below are storage claims, and they are on the list because the evidence is clean, not because they are likely to pay. No recovery amount on this page, or on any page, is guaranteed.
Every finding, line by line
Fourteen rows, six categories, one deadline each. This is the table, unedited — including the two rows valued at Amazon’s estimate rather than the seller’s cost, which is a self-inflicted loss and is marked as one.
| # | SKU / category | What went wrong | Evidence attached | Recoverable | Days left |
|---|---|---|---|---|---|
| F-01 | HG-TWL-BATH-6PK Fee overcharge — size tier |
Poly-bagged towel set scanned with air in the bag. Billed on 18.4 × 14.2 × 8.3 in / 4.1 lb — one dimension past the large-standard ceiling, so every unit bills a full tier up. | Measured at rest, 5 units: 17.6 × 13.5 × 7.4 in / 3.9 lb. Calipered photographs, supplier pack spec, and Amazon’s own recorded dimensions for contrast. | $1,244 | 44 |
| F-02 | HG-RUG-RUNNER-2X6 Fee overcharge — weight |
Rolled runner weighed damp on receipt after a humid inbound. Billed at 11.8 lb against a true 9.4 lb, crossing a shipping-weight band. Dimensions on file are correct; only the weight is wrong. | Measured 9.4 lb on a certified scale, 6 units, mean deviation 0.1 lb. Supplier weight certificate for the same production run. | $409 | 12 |
| F-03 | HG-NAPKIN-SET-4 Fee overcharge — size tier |
Cloth napkin set billed as large standard on 15.4 × 12.3 × 0.8 in / 1.1 lb. The per-unit error is $0.31, which is why nobody files it. The SKU ships about 640 units a month, so 1,910 of them sit inside the dispute window. | Measured 14.6 × 11.8 × 0.7 in / 0.9 lb — inside small standard on every axis. 8 units measured, packaging unchanged across three production runs. | $592 | 51 |
| F-04 | HG-STORAGE-BIN-L Fee overcharge — stale record |
Supplier moved to a slimmer carton two production runs ago. Amazon still bills the old measurement: 24.0 × 16.0 × 14.5 in. Nothing notified the seller, because nothing ever does. | Current production measures 22.5 × 15.5 × 13.0 in. Dated supplier carton spec, the changeover purchase order, and 4 units measured from current stock. | $926 | 29 |
| F-05 | HG-DUVET-QN-WHT Lost in the fulfillment center |
31 units received, then adjusted out of inventory over six weeks with no order, no return, no removal and no disposal record. They were simply subtracted. | Inventory ledger adjustments matched against the shipment receipt, plus a reconciliation showing no offsetting reimbursement in the payments ledger. Supplier invoice on file at $38.00/unit. | $1,178 | 22 |
| F-06 | HG-PILLOW-STD-2PK Lost in transfer |
64 units left one fulfillment center on an internal transfer. 51 arrived. The other 13 have not appeared in any facility for 71 days. | Transfer manifest against the receiving record — both Amazon’s own reports, disagreeing with each other. Supplier invoice on file at $16.45/unit. | $214 | 9 |
| F-07 | HG-THROW-KNIT-GRY Damaged in Amazon’s custody |
18 units moved to unfulfillable with a warehouse-damage disposition. Not carrier damage, not customer damage — damaged by the facility, which makes it Amazon’s. | Inventory adjustment detail with disposition codes, against the same units’ receiving record. No supplier invoice on file — valued at Amazon’s estimate, not cost. | $199 | 34 |
| F-08 | HG-SHEET-KING-IVY Return never returned |
22 customer refunds issued. 9 of those units never came back to inventory, are now past the return-to-inventory window, and were never reimbursed. | Returns report joined to the reimbursement ledger, order by order, showing 9 refunds with no matching return and no matching credit. Supplier invoice on file at $29.00/unit. | $261 | 17 |
| F-09 | HG-MAT-KITCHEN-ANTI Refund exceeded the order |
14 orders refunded at $46.99 against an order value of $38.99. Amazon returned the customer more money than the customer paid, and charged the difference to the seller. | Order detail against the refund line in the transaction ledger for each of the 14 order IDs. Pure arithmetic — no valuation dispute available to either side. | $112 | 27 |
| F-10 | HG-BASKET-WOVEN-M Return disposition wrong |
11 returns coded “customer damaged” and charged to the seller. Amazon’s own returns report records the same 11 units arriving at the facility intact and being damaged after receipt. | Returns report arrival condition against the final disposition, per unit, with the timestamps between them. Supplier invoice on file at $28.00/unit. | $308 | 6 |
| F-11 | HG-BIN-STACK-3PK Storage fee — wrong volume |
Monthly storage billed on 2.94 cu ft per unit for five consecutive months. Amazon’s own dimension record for the same SKU works out to 1.88 cu ft. | Storage fee report against the FBA dimension record, month by month. Amazon’s two systems disagreeing with each other, which is the strongest evidence there is. | $487 | 38 |
| F-12 | HG-LAMP-TABLE-CER Aged-inventory surcharge |
Aged-inventory surcharge assessed on 46 units that had already been removed and confirmed shipped out of the network four days before the snapshot date. | Removal order completion report with outbound tracking, set against the surcharge line and its snapshot date. Four days of daylight between them. | $203 | 19 |
| F-13 | HG-COASTER-SET-4 Inbound received short |
300 units shipped on one inbound. 276 received, and the shipment closed. The remaining 24 were never reconciled and never reimbursed. | Supplier invoice (what Amazon’s research form asks for) and the shipment reconciliation page showing 276 of 300 located; carrier proof of delivery and packing list held for the appeal. No supplier invoice on file — valued at Amazon’s estimate. | $100 | 31 |
| F-14 | HG-VASE-GLASS-TALL Removal never arrived |
Removal order for 40 units. 33 arrived at the return address. 7 were confirmed shipped by Amazon and never delivered anywhere. | Removal order detail with Amazon’s own carrier tracking showing no delivery scan, plus the seller’s dated receiving log. Supplier invoice on file at $40.60/unit. | $284 | 3 |
Sample data throughout. SKU codes, dimensions, unit costs and dollar amounts are invented for this page. Red days-left figures are the ones that would be worked first regardless of size — deadline outranks value.
The arithmetic behind one fee row
Every finding in the report carries its own working, because “$1,244 recoverable” with nothing behind it is a number you have no way to check. Here is F-01 opened up. This is the section our competitors’ reports do not have, and the reason is that most of them do not run dimension checks at all — the category was untouched by the March 2025 valuation change and it is still treated as a footnote.
And then the part that is worth more than the claim: this SKU ships about 160 units a month. Once the measurement on file is corrected, the seller stops paying $2.56 a unit forever — roughly $410 a month that never gets charged again. The report says so on the row, because a refund is a one-off and a corrected record is an annuity.
What we found and are not claiming
This page is in the report before the invoice is, and it is the reason we publish the sample at all. Anyone can send you a list of things they intend to bill you 10% of. The interesting page is the one that costs us money.
Amazon judges claim quality per account. A stream of weak, expired or duplicate filings buys you slower handling and more reflexive denials on the claims that are genuinely good. So the refusals are not modesty — they are the reason the fourteen above have a chance.
| What the reconciliation found | Face value | Why it is not on the findings list |
|---|---|---|
| 22 unresolved inventory adjustments across 9 SKUs, 71–140 days old | $1,430 | Expired. Outside the claim window, which makes them automatic denials rather than long shots. They are listed anyway so the seller can see exactly what waiting cost — that is the whole argument for daily reconciliation, in one line item. |
| Lost and damaged units on 14 SKUs already credited | $2,180 | Amazon already paid it, unprompted. Re-filing generates duplicates and would let us bill commission on money that was arriving anyway. We reconcile against reimbursements already issued before anything gets scored, which is why this number is on the refusal page and not in the headline. |
| 3 SKUs with recorded dimensions ~0.2 in off the measured mean | — | Real, and worth nothing. Inside normal measurement tolerance, and none of the three crosses a tier boundary, so the fee does not change. A discrepancy is not a claim. Logged and re-scored every cycle in case the packaging moves. |
| 6 SKUs with a credible dimension error and no sellable units in any facility | $2,940 | Unverifiable today. There is nothing in the network for Amazon to remeasure, so the request gets closed and the monthly remeasurement allowance is spent for nothing. On the re-file queue; it returns automatically the week a shipment is received. |
| 2 cases already open from the seller’s previous provider | $610 | Duplicate. A second case on the same event gets closed and makes the account look disorganised to the people deciding the rest of the claims. Left alone. If they are still open in three weeks we will ask before touching them. |
| 41 units disposed as non-compliant — missing suffocation warning on the poly bag | $486 | Genuinely the seller’s. Disposed under a policy the seller accepted, and no appeal makes that untrue. Fixed at the supplier instead, which is worth more than the $486 was. |
| 9 customer-damaged returns on two SKUs | $243 | The customer broke it, not the warehouse. The arrival condition and the disposition agree, which is exactly what makes F-10 above filable and these nine not. Same check, opposite answer. |
Sample data. “Face value” is what the finding would be worth if it were claimable, shown so the refusal is auditable rather than a shrug.
Add it up: $7,889 of things we found and will not file, against $6,517 we will. A service paid on volume has every reason not to show you that ratio. The full ranking logic, and the complete list of what we refuse on principle, is published at what we file and refuse.
One more, kept separate because it is not a refusal so much as a warning: two of the fourteen findings are storage claims, worth $690 between them. They are filed because the evidence is Amazon’s own records contradicting each other. They are also the category with the worst odds we have measured — 19% of the storage claims we filed on our own account were paid. The report says that on the row rather than in a footnote, because you should discount that $690 heavily before you plan around it.
Two missing invoices, $252 gone
Since March 2025, lost and damaged inventory reimburses at your manufacturing cost rather than sale price. The part that quietly costs sellers money is the second half of that rule: if you have not given Amazon a cost for a SKU, it applies its own internal estimate, and that estimate runs below what sellers actually pay. The full mechanics are here.
So every inventory-valued finding in this report carries a documentation status, and the report shows the difference in dollars rather than describing it. On this sample account, seven of the fourteen findings are valued at cost. Five have a supplier invoice on file. Two do not.
| Finding | Cost on file? | Per unit used | Actual cost | Difference |
|---|---|---|---|---|
| F-05 · HG-DUVET-QN-WHT — 31 units | Yes — supplier invoice | $38.00 | $38.00 | — |
| F-06 · HG-PILLOW-STD-2PK — 13 units | Yes — supplier invoice | $16.45 | $16.45 | — |
| F-08 · HG-SHEET-KING-IVY — 9 units | Yes — supplier invoice | $29.00 | $29.00 | — |
| F-10 · HG-BASKET-WOVEN-M — 11 units | Yes — supplier invoice | $28.00 | $28.00 | — |
| F-14 · HG-VASE-GLASS-TALL — 7 units | Yes — supplier invoice | $40.60 | $40.60 | — |
| F-07 · HG-THROW-KNIT-GRY — 18 units | No — Amazon’s estimate | $11.05 | $20.70 | −$174 |
| F-13 · HG-COASTER-SET-4 — 24 units | No — Amazon’s estimate | $4.15 | $7.40 | −$78 |
Sample figures. Amazon’s estimates here are invented at roughly half of actual cost, which is in the range we see, not a published ratio.
Two invoices, on two ordinary SKUs, and 46% of that money evaporates. Now apply the same ratio to a catalog where nothing is documented, which is the state most accounts are in, because nothing in Seller Central ever prompted anyone to fill that field in.
Which is why the fix comes before the filing. We collect supplier invoices during onboarding and get costs on file first, then file. Filing fast on an undocumented SKU locks in the lower number permanently — it is not something you can appeal your way out of afterwards. A week’s delay to get the paperwork right is worth more than a week’s head start, and a service that files everything immediately is optimising for its own activity report.
The documentation status column stays in every monthly report from then on, so a SKU that arrives undocumented does not stay that way quietly. Reference: lost & damaged inventory and the policy tracker.
Now get one with your numbers in it.
The audit is free, it takes about ten minutes to set up, and it produces this report on your catalog — refusals, expiry clock, documentation gaps and all. You keep it whether or not you hire us, and you are welcome to take it to whoever you are using now and ask why they did not find it.
Everything on this page is sample data on a fictional seller. Recovery outcomes are decided by Amazon and no specific amount is guaranteed — see our own published results, including what did not pay.