> For the complete documentation index, see [llms.txt](https://docs.sellermagnet.com/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.sellermagnet.com/inventory-and-operations/fee-audit.md). # Fee Audit Audit the fees Amazon charged on your settled orders, verify every flagged overcharge against Amazon's own published rates before you file, and recover the difference on the claims that are genuinely {% hint style="info" %} **Difficulty:** ๐ŸŸก Intermediate ยท **Reading time:** \~15 min {% endhint %} {% embed url="" %} ## ๐Ÿ“‹ Overview Amazon charges dozens of different fees across referral fees, FBA fulfillment fees, closing fees, and other per-item charges. Sometimes a fee looks wrong: a category rate misapplied, an FBA size tier measured incorrectly, or a charge that does not match Amazon's own schedule. The **Fee Audit** tool scans the fees Amazon actually charged on your settled orders, flags anything that looks off, and then **verifies each flag against Amazon's own published rates before it ever tells you to file a claim**. {% hint style="success" %} **Flag first, then verify.** Fee Audit works in two stages. The scan flags any fee more than 5% above Amazon's own fee estimate for that product. A separate verification engine then re-derives what Amazon should have charged from your own settled orders and reaches a verdict on each flag. Money (Actual, Expected, Overcharge, Total) is shown **only where that verdict proves the row is genuinely claimable** - everywhere else you see a dash and the reason we could not verify it. This is deliberate: filing a claim on a fee Amazon charged correctly sends you to Seller Support with a number Amazon refutes from its own ledger in one query, and that costs you credibility on the cases that are real. {% endhint %} {% hint style="warning" %} **Fee Audit requires an analytics plan.** The page is plan-gated. On an account without an analytics dashboard plan the whole module renders blurred behind an **"Upgrade Required"** overlay with a **View Plans** button. Once you are on a qualifying plan, the module itself carries no extra charge - there is no per-scan or per-claim fee. {% endhint %} *** ## ๐Ÿ“Š Key Performance Indicators Four tiles sit across the top of the page. The numbers are **verdict-driven**, not raw counts of flagged rows - the sub-line under each tile tells you what the verification engine actually concluded. | KPI | What the tile shows | Sub-line | | ---------------- | --------------------------------------------------------- | ----------------------------------------------------------------------- | | **Total Issues** | How many fees were flagged and checked | Verdict breakdown, e.g. *"Charged correctly: 194 ยท Not verifiable: 32"* | | **Claimable** | How many flags the engine proved are genuinely disputable | The claimable amount, or *"Nothing claimable"* | | **Claims Filed** | How many claims you have recorded | Money is deliberately shown as **"-"** (see note below) | | **Resolved** | How many claims closed | The amount **actually reimbursed** by Amazon | {% hint style="info" %} **Why "Claims Filed" shows a dash for money.** The value of a filed claim is not known until Amazon responds, and quoting the flagged figure would put a number on screen that has not been recovered. Only the **Resolved** tile shows money, and it shows the amount Amazon *actually reimbursed* - which you enter yourself when you mark an item resolved (Amazon often reimburses part of a claim, not the full amount). {% endhint %} > **Quick win:** Start with the **Claimable** tile. That count comes straight from the verification engine - these are the flags where Amazon's own published rate demonstrably differs from what it charged, with enough evidence behind them to defend at Seller Support. Everything else is shown for transparency, not because it is actionable. *** ## ๐Ÿงฎ The verdict engine This is the heart of Fee Audit and the reason its numbers are trustworthy. Every flagged row is adjudicated and carries one of five verdicts, shown in the **Recommendation** column and in the detail panel: | Verdict | Meaning | | ------------------------- | ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- | | **Charged correctly** | Amazon's published rate for this product reproduces the fee it charged. No overcharge. This is the common, and valuable, result - proof you are being billed right. | | **Charged below rate** | Amazon charged *less* than its published rate. Nothing to reclaim, but worth watching in case the rate changes. | | **Discrepancy verified** | Amazon charged more than its own published rate, the model reproduces your account's fees, and there are enough units behind it. This is the only verdict that is claimable. | | **Not verifiable** | We cannot reproduce Amazon's fee arithmetic for this marketplace accurately enough to judge the row (unsolved marketplace, a fee model that does not match your own recent fees, a line carrying shipping/gift wrap, or a sub-floor per-unit price). Shown, but not actionable. | | **Difference not proven** | A discrepancy exists but we deliberately will not stand behind it - too few units, or a fitted rate that is not an actual published Amazon rate. No claim to file. | {% hint style="success" %} **Money fails closed.** The engine reproduces Amazon's two-step fee computation (rate on the correct base, then fee VAT) to the **cent**, using your own settled orders rather than a generic fee table. If it cannot reproduce that arithmetic, it reports **"-"** rather than guessing. That is why most rows show a dash: the honest answer is usually "this fee is correct" or "we cannot prove otherwise," and the tool says so instead of inventing a recoverable figure. {% endhint %} *** ## ๐Ÿ“ˆ Charts * **Findings by Fee Type** (Bar Chart): how **many** findings sit in each fee type (Referral Fee, FBA Fulfillment Fee, Closing Fee, Other Fee). This chart plots **counts, not money** - a count has no currency and needs no verdict, so it is safe to read at a glance. * **Issue Distribution** (Pie Chart): the share of findings across the fee types. > **Note:** These charts count findings, not euros. A tall bar means many fees of that type were flagged for checking - not that a large sum is recoverable. The Claimable KPI and the Recommendation column tell you what is actually disputable. *** ## Audit Types Fee Audit derives the type of each finding from the fee Amazon charged. The **Audit Type** filter offers four options: | Type | What it covers | How it is checked | | ----------------------- | ----------------------------------------------------------------------- | ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------ | | **Referral Fee** | Amazon's category commission on each sale | Re-derived against Amazon's published referral rate on the correct fee base. This is the type that can become claimable. | | **FBA Fulfillment Fee** | The flat fulfillment charge for the product's size tier | A flat charge, **not** a percentage, so it is never verified against a rate. A gap usually means Amazon's measured dimensions/weight differ from your listing - the fix is a re-measure, not a rate dispute. | | **Closing Fee** | The fixed closing fee on media categories | A fixed amount per category, not a percentage, so it is not rate-verified. Check the listing's category before opening a case. | | **Other Fee** | Any other per-item fee Amazon charged that came back above the estimate | Flagged for visibility; generally **Not verifiable** because there is no published rate to reproduce. | {% hint style="info" %} **Legacy types.** Older audit rows may still carry the labels **Wrong Category** or **Weight / Dimension**. These are from an earlier version of the scan and are shown only on those historical rows - they are **not** options in the Audit Type filter. New findings are always one of the four types above. {% endhint %} *** ## ๐Ÿ‘ฃ Step-by-Step Example: A verified referral-fee discrepancy > **Scenario:** You sell a homeware product on Amazon.de. The verification engine has fitted your account's referral behaviour from your own settled orders and can reproduce Amazon's fees to the cent on most of them. On one ASIN, Amazon charged a referral fee that is higher than its own published rate for the product's category on the correct (VAT-inclusive) base - and it did so across enough units to be defensible. > > **What the row shows:** > > **What to do:** > > 1. Click the row to open the detail panel and read **"Why this is on your list"** - the *Flagged by the scan* line and the *What we found when we checked it* verdict. > 2. Follow the referral-rate guidance block: **Seller Central > Help > Get Support > Selling on Amazon > Your account > Fees and charges**. > 3. Click **File Claim**, copy the auto-generated template, and paste it into the case. Attach the sample order's detail page. > 4. Keep it to one product per case, and quote Amazon's published rate. | Field | Value | | ----------------- | ---------------------------------------------------- | | Audit Type | Referral Fee | | Recommendation | **Discrepancy verified** - *Open a case with Amazon* | | Actual Fee | (the fee Amazon actually billed) | | Expected Fee | (Amazon's published rate on the correct base) | | Overcharge / Unit | the per-unit gap | | Total Overcharge | the whole-order-item gap, per the evidence | | Status | Open | {% hint style="success" %} **This is the exception, not the rule.** Across a real fleet audit, the overwhelming majority of flagged fees adjudicated to **Charged correctly** or **Not verifiable**. A genuinely verified, claimable discrepancy is uncommon - which is exactly why the tool suppresses money everywhere else. When you do get one, it is worth filing precisely because it is defensible. {% endhint %} *** ## ๐Ÿ‘ฃ Step-by-Step Example: A flag that could not be verified > **Scenario:** The scan flagged an FBA Fulfillment Fee that came back higher than Amazon's estimate. When the engine checked it, the row landed on **Not verifiable** - FBA fees are a flat size-tier charge, not a percentage, so there is no rate to reproduce. > > **What the row shows:** > > **What the detail panel tells you:** the FBA guidance block explains this is a **size and weight question, not a rate question**. A difference here usually means Amazon's measured dimensions or weight differ from your listing. > > **What to do (outside a fee claim):** > > 1. **Seller Central > Reports > Fulfilment > Fee Preview** - check the size tier and dimensions Amazon holds. > 2. Compare them with the actual packed unit. If they differ, measure and weigh it yourself. > 3. **Seller Central > Help > Get Support > Fulfilment by Amazon > FBA fee or product-size issue** - request a re-measure and attach your measurements and a photo next to a ruler. > 4. If the re-measure agrees with you, ask for the difference to be refunded for the affected period. | Field | Value | | -------------------------------------- | ------------------------ | | Audit Type | FBA Fulfillment Fee | | Recommendation | **Not verifiable** | | Actual / Expected / Overcharge / Total | **-** (money suppressed) | | Status | No action needed | {% hint style="warning" %} **A dash is information, not a failure.** When a row shows "-" and "Not verifiable," it does not mean the scan broke. It means the tool checked the fee and could not prove an overcharge - so it declines to hand you a number you cannot defend. Open the detail panel to see exactly what was checked and why. {% endhint %} *** ## Audit Items Table ![Fee Audit dashboard - Overcharge detection and reimbursement claims screenshot](https://4139654583-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2Fih2knSvVpBcq5rIeXVjM%2Fuploads%2Fgit-blob-204889ca1193558029f12be77d214b1d4c1cdf7a%2Ffee-audit.png?alt=media) | Column | Description | | ---------------------------- | ------------------------------------------------------------------------------------------------------------------- | | Checkbox | Select for bulk actions | | **Marketplace** | Country flag and name | | **Audit Type** | Fee type badge | | **Product** | Product thumbnail and name | | **ASIN** | Amazon Standard Identification Number (links to Amazon) | | **SKU (Stock Keeping Unit)** | Seller SKU | | **Actual Fee** | What Amazon charged - shown only where a verdict backs it, otherwise **-** | | **Expected Fee** | Amazon's published rate on the correct base - shown only on claimable rows, otherwise **-** | | **Overcharge** | Per-unit gap - **-** unless the row is claimable | | **Occurrences** | Units behind the finding | | **Total Overcharge** | The whole gap - **-** unless the row is claimable | | **Recommendation** | The verdict badge (Charged correctly, Not verifiable, Discrepancy verified, and so on) plus the suggested next step | | **Status** | Open, Claim Filed, Resolved, Dismissed (or "No action needed" where the verdict overrides) | | **Detected** | When the finding was recorded | | **Actions** | View Details, and File Claim **only** where a claim can actually be built | {% hint style="info" %} **Rows charged correctly are filtered out of the list.** They are counted in the KPI sub-line but not printed as rows, so the table stays focused on what needs a look. When that happens the pagination line notes it, e.g. *"194 charged correctly, not listed."* {% endhint %} The table also carries a **Show dismissed** checkbox (on the **All** tab), a per-page selector of **10 / 25 / 50 / 100** rows, sortable columns, and a search box that matches ASIN, SKU, and fee type. *** ## Filing Claims ![Reimbursements module: track lost and damaged inventory claims screenshot](https://4139654583-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2Fih2knSvVpBcq5rIeXVjM%2Fuploads%2Fgit-blob-8cd16dccf0e1320ac6e893b5094dcac841f1d15c%2Freimbursements.png?alt=media) {% hint style="info" icon="circle-info" %} **Filing records the claim in SellerMagnet; it does NOT auto-submit to Amazon.** SellerMagnet tracks the claim's status and hands you a ready-to-paste template, but you still open the actual case in Seller Central yourself. The optional Case ID field is where you link the two together. {% endhint %} The **File Claim** action appears only on rows the engine proved claimable (verdict **Discrepancy verified**). Clicking it - from the row or the detail panel - opens the **File Fee Overcharge Claim** modal, which has two optional inputs plus a ready-to-send template: | Field | Notes | | ------------------ | ---------------------------------------------------------------------------------------------------------------------------------- | | **Amazon Case ID** | Optional. Fill this in if you have already opened a case in Seller Central (e.g. `1234567890`) so you can cross-reference it later | | **Notes** | Optional. Free-text notes kept with the claim for your own records | ### Claim summary and template The modal shows a **claim summary card** for the item you are filing - the product name, the audit type, and the total overcharge - so you can confirm you are filing against the right issue. Beneath it is a pre-filled **claim template** (*"Claim Template (copy & paste into Seller Central)"*). SellerMagnet assembles a complete support message from the item's verified evidence: the fee type, ASIN/SKU, the fee Amazon charged, the expected fee at the published rate, the per-unit overcharge, the number of units, the total overcharge, and a sample order ID. Use **Copy to Clipboard** to grab the text and paste it into a Seller Central support case. Click **Submit Claim** to record the claim in SellerMagnet (this moves the item to **Claim Filed**). > **Pro tip:** You do not have to write the claim from scratch. SellerMagnet auto-generates the full template from the item's own verified numbers. Just hit **Copy to Clipboard** and paste it into your Seller Central case. ### Dismissing an issue To set an issue aside, use **Dismiss** (from the row, the detail panel, or in bulk). The **Dismiss Issue** modal lets you add an optional **reason** and notes that the action is reversible - dismissed items move to the **Dismissed** tab and can be reopened later. ### Reopening and resolving * **Reopen** (detail panel, on a dismissed item): puts a dismissed finding back into the workflow. Useful because the audit also auto-dismisses findings a new scan no longer detects, and you may want one back. * **Mark as resolved** (detail panel, on a filed claim): closes the loop. It prompts you for **how much Amazon actually reimbursed** - leave it blank if you do not know. That figure, not the claimed amount, is what feeds the **Resolved** KPI, so the tile reports real recovered money. ### Bulk actions Tick the checkboxes on multiple rows to reveal the bulk action bar, then choose: * **File Claims** - marks every selected claimable item as **Claim Filed** in one step. Non-claimable selections are skipped, with a note. * **Dismiss** - marks every selected item as **Dismissed** in one step. ### Export Use the **Export** dropdown to download the current view as **CSV** or **Excel** for your records or your accountant. The export honours your current filters and tab. *** ## Claim Status Lifecycle Every audit item carries a stored status that tracks where it sits in the recovery workflow. Note that the **Recommendation** verdict is a separate question - a row can be stored as "Open" while its verdict reads "No action needed," in which case the status badge shows **No action needed** to avoid contradicting itself. | Status | Meaning | | --------------- | --------------------------------------------------------------------------------------------------------------------------------------------------------------------------- | | **Open** | The finding has not been triaged yet. Every flagged row starts here. (This is the tab labelled "Open"; the underlying stored status is the workflow's action-needed state.) | | **Claim Filed** | You have recorded a claim (individually or in bulk) and are awaiting Amazon's response | | **Resolved** | You marked the claim closed and recorded what Amazon reimbursed - the item shows no further actions | | **Dismissed** | You set the item aside; reversible via **Reopen** | {% hint style="info" %} **Legacy status.** Some historical rows may still show **Detected** from an earlier version. New findings do not use it. {% endhint %} *** ## Fee Audit Review Process Template Use this structured process to get the most out of Fee Audit: ### Weekly Review (10 minutes) * [ ] Open the Fee Audit page and check the **Claimable** KPI. * [ ] Switch to the **Open** tab and read the **Recommendation** column. * [ ] File claims for any **Discrepancy verified** rows, highest total first. * [ ] Dismiss anything you have reviewed and decided not to pursue. ### Monthly Deep Review (30 minutes) * [ ] Set the **Audit Window** and click **Run Audit** for a fresh scan. * [ ] Review the verdict mix in the **Total Issues** sub-line - mostly "Charged correctly" is a healthy account. * [ ] For any **Not verifiable** FBA rows, follow the re-measure guidance in the detail panel. * [ ] Check the **Resolved** tab and confirm Amazon's reimbursements match what you recorded. * [ ] Follow up on any **Claim Filed** items older than 30 days, then **Mark as resolved** with the actual amount. * [ ] Export a monthly report for your financial records. ### Quarterly Strategic Review (45 minutes) * [ ] Run a fresh audit across a longer window (180 or 365 days). * [ ] Look at which fee types recur in the **Findings by Fee Type** chart. * [ ] For products with recurring FBA size-tier gaps, request cubiscan re-measures in Seller Central. * [ ] Review Amazon's latest fee schedule for changes that may affect your catalog. *** ## ๐Ÿ” Filters | Filter | Options | | ---------------- | ---------------------------------------------------------------------------------- | | **Marketplace** | All Marketplaces, or a specific connected marketplace | | **Audit Type** | All Types, Referral Fee, FBA Fulfillment Fee, Closing Fee, Other Fee | | **Status** | All Statuses, Action Needed, Claim Filed, Resolved, Dismissed | | **Audit Window** | Last 30, 60, 90 (default), 180, or 365 days - applied when you press **Run Audit** | Use **Apply** to run the filters and **Reset** to clear them. *** ## Detail Side Panel Click any row to open the detail panel. It gives you the full evidence behind the finding: * **Why this is on your list** - two plainly stated halves: *Flagged by the scan* (the fee was more than 5% above Amazon's estimate when the audit ran) and *What we found when we checked it* (the verdict badge and its explanation). * **Fee Comparison** - Actual, Expected, per-unit and total overcharge. Money is shown only where a verdict backs it; on non-claimable rows the panel deliberately does not print an Expected Fee you could subtract and act on. * **Additional Info** - detected date, sample order, case ID, marketplace, and (once closed) the resolved amount. * **Context-specific guidance** - a block tailored to the case: **FBA re-measure** steps, a **referral-rate dispute** walkthrough, a **closing-fee** category note, or a **"No case to open"** explanation where nothing is provable. * **History** - the full audit trail of status changes and notes. * **Footer actions** - Dismiss, File Claim (claimable rows only), **Reopen** (dismissed rows), and **Mark as resolved** (filed claims). > **Best practice:** Always open the detail panel before acting. The "Why this is on your list" section and the guidance block tell you whether you have a real case and exactly where to raise it - which keeps you from opening thin cases that erode your credibility with Seller Support. *** ## Running an Audit Before you scan, set the **Audit Window** in the filter bar - how far back to look at your settled orders: **Last 30, 60, 90 (default), 180, or 365 days**. A longer window catches more history but takes longer to process. Click **Run Audit** to trigger a new scan. The button appears in the table toolbar (as the page's primary action) and again next to the **Last audit** timestamp at the bottom of the page. Each run also **auto-dismisses** open findings the new scan no longer detects, so the list always reflects the latest audit. {% hint style="info" %} **First-run auto audit.** Because the audit is opt-in, a brand-new account lands on an empty page. To avoid a blank screen with no explanation, SellerMagnet **automatically starts one scan** for an account that has never been audited. It is one-shot and rate-limited (roughly a 24-hour cooldown per account), so it never loops - and the manual **Run Audit** button is unaffected. {% endhint %} {% hint style="warning" %} **Allow processing time.** A full fee audit can take several minutes and, on large catalogs, longer - the scan calls Amazon's fee-estimate API once per product per marketplace at about one request per second, and the verification engine re-derives every fee from your settled orders. A banner shows the run is in progress; it keeps running in the background even if you navigate away, so you can refresh later to see the results. {% endhint %} *** ## ๐Ÿ”„ Before & After: What Fee Audit changes | Metric | Before SellerMagnet | After SellerMagnet | | ----------------------------- | -------------------------------------------------------------- | ---------------------------------------------------------------- | | Fee verification | Manual spot checks, if any | Every settled-order fee scanned and adjudicated | | Confidence in a claim | Guesswork against a generic fee table | Verified against Amazon's own published rate on your own orders | | False claims filed | Common - a discrepancy vs an estimate looks like an overcharge | Suppressed - money only appears where it is provable | | Time spent on fee review | Hours, or never done | Minutes per week | | Knowing your fees are correct | Unknown | Stated explicitly - "Charged correctly" is a result, not silence | *** ## Seasonal & Timing Advice {% hint style="danger" %} **January and June: Amazon's fee change windows.** Amazon typically updates its fee schedules in January (effective February) and June (effective July). Run a complete audit within 2 weeks of any fee change - new tiers or rates may be misapplied to your products during the transition. {% endhint %} {% hint style="warning" %} **After listing updates:** If you update a product's weight, dimensions, or category, run a fee audit 7-14 days later. Amazon's systems may not immediately reflect the change, and you could be charged incorrectly during the transition. {% endhint %} > **Q4 volume check:** During Q4 your order volume rises, so any genuine fee error is multiplied across more transactions. Run audits more often through October to December, and act on the verified findings promptly - Amazon's claim windows are limited. {% hint style="info" %} **New product launches:** When you launch a new product, Amazon assigns initial weight, dimensions, and category during the first FBA receiving. Run an audit within 2-3 weeks of your first FBA shipment being received - this is when classification errors most commonly appear, and they show up as FBA size-tier gaps you can raise for a re-measure. {% endhint %} ***
โš ๏ธ Common Mistakes to Avoid {% hint style="danger" %} * **Mistake 1: Treating every finding as an overcharge.** A flagged row is a *candidate*, not a proven claim. Read the Recommendation verdict before you act - most flags adjudicate to "Charged correctly" or "Not verifiable." * **Mistake 2: Filing on a "-" row.** If money is suppressed, the tool could not prove an overcharge. Filing anyway sends Amazon a number it refutes from its own ledger and weakens your credibility on the real cases. * **Mistake 3: Ignoring FBA "Not verifiable" rows.** These are not dead ends - they are usually a size-tier measurement problem. Follow the re-measure guidance in the detail panel; the fix lives in a Fulfilment case, not a fee dispute. * **Mistake 4: Recording the claimed amount as the resolved amount.** Amazon frequently reimburses part of a claim. Enter what Amazon *actually* paid when you Mark as resolved, so the Resolved KPI stays honest. * **Mistake 5: Not re-running after fee schedule changes.** Products correctly classified before a change can be misclassified afterward. A fresh Run Audit re-checks everything against the current rates. {% endhint %}
## ๐Ÿ”ง Troubleshooting ### Why does the audit show zero findings? 1. **Your fees may genuinely be correct.** Most accounts, most of the time, are billed correctly - and the tool says so rather than inventing issues. 2. **Run a fresh audit.** Click **Run Audit** and wait for it to complete. The **Last audit** timestamp should update. 3. **Check your filters.** You might be viewing a marketplace, type, or status with no matching rows, or a window that is too short. 4. **Verify your account connection.** Go to Settings > Connections and confirm your Amazon SP-API (Selling Partner API) access is active with the required permissions. ### Why is money shown as "-" on almost every row? Because the verification engine only shows money it can prove. It reproduces Amazon's fee arithmetic to the cent from your own settled orders; where it cannot (an unsolved marketplace, a model that does not match your recent fees, a line carrying shipping or gift wrap, a sub-floor price, or too few units), it reports a dash and a verdict such as **Not verifiable** or **Difference not proven** rather than a figure you might act on. Open the detail panel to see exactly what was checked. ### The "Expected Fee" is blank - how is it calculated? The Expected Fee is Amazon's published referral rate applied to the correct fee base for the marketplace (VAT-inclusive item price on EU/UK, principal on US/CA, and so on), reproduced from your own orders. It is shown **only** on rows the engine proved claimable. On every other row it is suppressed, because printing an "expected" we do not stand behind would invite you to subtract it from the actual fee and file on the difference. ### An FBA fee looks wrong but the verdict says "Not verifiable." FBA fulfillment fees are a flat charge for the size tier Amazon measured - not a percentage - so there is no rate to reproduce. A gap almost always means Amazon's recorded dimensions or weight differ from your product. Follow the FBA guidance block in the detail panel: check the Fee Preview, measure the packed unit yourself, and open a Fulfilment case requesting a re-measure. ### Why do some rows show "No action needed" as their status? The stored workflow status starts as "Open" on every flag, but the verdict may conclude there is nothing to do. Rather than show a red "Action Needed" next to a verdict that says otherwise, the status badge reads **No action needed** so the row does not contradict itself. *** ## โ“ FAQ
How far back can I claim fee overcharges? Amazon's policy generally allows claims for fee errors going back up to 90 days for most fee types; some categories differ. File promptly - the sooner you claim a verified discrepancy, the more likely it is to be approved.
What is a cubiscan and how do I request one? A cubiscan is Amazon's official process for remeasuring your product's weight and dimensions at the fulfillment center. Request one by opening a case in Seller Central under FBA support (product-size / FBA fee issue). Amazon will remeasure at the next opportunity and update the recorded dimensions. This is the right route for FBA "Not verifiable" findings.
Does SellerMagnet check fees across all my marketplaces? Yes. Fee Audit scans all connected Amazon marketplaces. Each marketplace has its own fee schedule and fee base, so the verification engine fits each one separately. A marketplace whose fee formula it cannot solve fails closed to "Not verifiable" rather than producing a guess. Use the Marketplace filter to review each one.
Why does the tool tell me a fee is correct instead of finding money? Because proving a fee is right is a genuine result. The verification engine is built to fail closed: it would rather tell you "Charged correctly" than send you to Seller Support with a number Amazon can refute. When it does surface a claimable discrepancy, that discrepancy is defensible precisely because everything unprovable was filtered out.
What if my product is right on the boundary of a size tier? Size tier boundaries are strict, so even a 1cm or a few-gram difference can change the fee tier. These show up as FBA findings. Verify the packed unit against the dimensions Amazon holds, request a re-measure if they differ, and consider trimming packaging to sit firmly in the cheaper tier.
How does Fee Audit differ from the Reimbursements tool? Reimbursements focuses on **lost, damaged, or missing inventory** - physical discrepancies in your stock. Fee Audit focuses on **financial overcharges** - Amazon charging the wrong fee amount. Both recover money, but they address different types of errors.
Do I need a paid plan to use Fee Audit? Fee Audit requires an analytics dashboard plan. Without one, the page is blurred behind an "Upgrade Required" overlay. On a qualifying plan the module carries no additional per-scan or per-claim charge.
How often should I run a fee audit? At minimum monthly, and more often during high-volume periods (Q4) or immediately after Amazon announces fee schedule changes. Because each **Run Audit** re-checks everything and auto-dismisses findings that no longer apply, running it regularly keeps the list current.
*** ## ๐Ÿ’ก Tips > **Pro tip:** Let the **Recommendation** column do the triage. File the "Discrepancy verified" rows, action the "Not verifiable" FBA rows through a re-measure case, and leave the "Charged correctly" rows alone - they are proof, not a problem. > **Weight & dimension gaps** are among the most common FBA findings. They surface as "Not verifiable" because they are not a rate question - request a **cubiscan re-measure** in Seller Central to correct the recorded dimensions. > **Pro tip:** Keep a record of your actual product weights and dimensions. When a finding points at a size-tier gap, you can attach your measurements to the re-measure case, which significantly improves the outcome. > **Pro tip:** When you Mark as resolved, enter the amount Amazon actually reimbursed. The Resolved KPI is the one money figure the module reports as real - keeping it accurate makes it worth trusting. *** ## โžก๏ธ What's Next? {% content-ref url="/pages/GQTlLpp6bu8z9YBUMb55" %} [Reimbursements (Lost & Found)](/inventory-and-operations/reimbursements.md) {% endcontent-ref %} {% content-ref url="/pages/7nzjHrzcKyOjaN4wV9pT" %} [Inventory Planner](/inventory-and-operations/inventory-management.md) {% endcontent-ref %}