> ## Documentation Index
> Fetch the complete documentation index at: https://sims-bae272dd.mintlify.site/llms.txt
> Use this file to discover all available pages before exploring further.

# Counts and discrepancies

> Settling the difference between what the system says and what is on the shelf.

**Stock & Inventory → B2B Reconciliation** compares expected stock against what
you actually counted, and holds the difference open until somebody explains it.

Three tiles: **Open Discrepancies**, **Total Unresolved Value** and **Resolved
This Month**. Filter by **All / Open / Investigated / Resolved**.

## Run a count

**1. Choose New Count**

Opens **New Physical Count**.

**2. Type what you actually counted**

Enter the real quantity for each product in the list.

**3. Add notes and create the records**

**Create Reconciliation Records** opens one record per line that did not
match.

## Where discrepancies come from

Counts are one source. The other is receiving: **any delivery line where the
counted quantity does not match the invoice opens a discrepancy automatically**,
so a short shipment cannot quietly become your problem later.

## Resolve one

Open the record and record three things:

**1. How you investigated**

Recounted the stock, reviewed invoices, reviewed POS sales, checked camera
footage, contacted the supplier, checked the backroom, or other.

**2. What it turned out to be**

Supplier short-shipped, damaged in transit, miscount, data entry error,
theft suspected, misplaced stock, supplier credit issued, write-off, or
other.

**3. Your conclusion, in writing**

A real sentence. A minimum length is enforced — "ok" is rejected — because
the value of this record is entirely in what it tells the next person.

> [!WARNING]
>
> A resolved record is permanent. If the conclusion turns out to be wrong, a new
> record is created that points back at it rather than the old one being edited.

## Getting value out of it

Look at the **resolution categories** over a few months rather than one record
at a time. A supplier who is repeatedly short by a case or two, or one shift
that produces most of your "theft suspected" entries, is only visible in the
pattern.


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.