Amazon Inventory Ledger Summary vs Detailed View: Which Report Should You Use?

The two views answer different questions

Amazon describes Summary View as a reconciliation of starting inventory, movements and ending balance. Detailed View exposes individual movements, including dates, product identifiers, references, quantities and reconciliation fields. Those report definitions explain why a balance mismatch and a potential reimbursement event need different investigations.

Our recommendation is to begin with the question you want answered. If you are checking a month's stock movement, start with the grouped view. If you are tracing a specific adjustment, use the event view. One can support the other, but neither by itself proves reimbursement eligibility.

Keep dates, locations and product scope aligned

A comparison becomes misleading when its two sides cover different populations. A country-wide monthly balance should not be compared with a single fulfillment center's daily movement and then labeled an unexplained loss. Write the filters beside your working table before doing any arithmetic.

Also check whether one file covers a single product identifier while another covers a wider set. Your working notes should state which products and locations are included. If the export settings are unclear, regenerate a small, known scope and use it to confirm your interpretation.

• Use the same reporting start and end dates for the initial comparison.
• Record the location grouping and product filters.
• Separate marketplace and account data.
• Retain the original quantity signs and report headings.

Worked example: a balance check becomes an event investigation

Consider a fictional product with 100 starting units, 20 units received and 30 units shipped to customers. Before other movements, the expected balance would be 90. If the grouped view ends at 87, the three-unit difference tells you where to look; it does not tell you whether those units were lost.

Suppose the event file then shows a three-unit movement for the product. Investigate its meaning, location and reference. It might describe an adjustment, a disposal or part of a wider sequence. Check subsequent activity and supporting records before changing its status to unresolved loss.

Keep the arithmetic as a diagnostic, with each component visible. Avoid adjusting the equation until the answer equals the displayed balance without recording which missing movement explained the difference. That would hide the very issue you are trying to resolve.

Preserve the columns that make an event traceable

When preparing a working copy, keep enough information to return to the original row. For this review, we recommend retaining date, identifiers, reference, location, quantity and reason or disposition information wherever supplied. Add your own review notes in new columns.

Do not rename a source column to estimated loss if it actually contains a movement quantity. Source data and your interpretation serve different purposes. The same principle applies to reconciled and unreconciled quantities: investigate them in context instead of using one number as an automatic claim amount.

• What an unreconciled quantity actually tells you
• Prepare CSV and Excel exports without changing identifiers

Avoid three common comparison mistakes

First, do not add overlapping periods as if they were independent activity. A full-month file and a file for the month's last week can contain the same events. Keep an import manifest and distinguish a new export from new inventory activity.

Second, avoid using product names as the only join key. Third, check whether a later report has updated the interpretation of an earlier event. Preserve the newest evidence alongside the original observation so another reviewer can understand why the result changed.

Use the right export in FluxPilot

For an event investigation in FluxPilot Recover, begin with the detailed ledger export and retain the source fields. Optional reimbursement history and cost data help you review coverage and value after the inventory candidates have been identified. A summary-only file may lack the event detail needed for that work.

If a result looks unexpectedly empty or broad, verify the source file and filters before assuming there are no issues or that every listed row is actionable. A small, known example is a better starting check than uploading multiple incompatible exports at once.