A physical inventory count is the process of manually counting and checking assets or stock on hand to confirm that records match what actually exists.
A physical inventory count is the process of going to where assets or stock are supposed to be and counting what is actually there, so that records can be checked against reality. It is the floor-side half of keeping a register honest: the count establishes what exists, and asset reconciliation deals with any differences between that and what the records claim.
How a count works
The shape is the same whether you are counting desktop computers or boxes of gloves. Define the scope - a location, a category, or everything. Freeze movement as far as practical, because counting items that are mid-transfer guarantees phantom discrepancies. Walk the space, identify each item positively - by scanning its tag, not by recognising its silhouette - and tick it off against the expected list. Note three kinds of exception: items missing, items present but not on the list, and items present but in the wrong place or visibly damaged. Finish with a signed summary of what was counted, by whom, and when.
Stricter counts are done blind: the counter records what they find without seeing the expected quantities first, so the list cannot suggest the answer.
Cycle counts vs wall-to-wall counts
A full (wall-to-wall) count covers everything in scope in one exercise. It gives a complete snapshot - which is why year-end accounts and a fixed asset audit often demand one - but it is disruptive and tends to be rushed.
A cycle count spreads the work out: a subset is counted each week or month on a rotation, so the whole estate is covered over a period without ever shutting anything down. Sensible rotations weight by risk - portable, valuable, frequently-moved items get counted more often than the bolted-down ones.
Most teams land on a blend: cycle counts through the year, one full count to anchor the books.
How scanning speeds it up
The slow part of a count is not walking - it is identifying each item and matching it to a record. Reading a label and typing an ID into a list invites transcription errors and takes the better part of a minute per item; scanning a barcode or QR label takes seconds and matches the record exactly. Scans also produce their own timestamped evidence of what was checked and when, which is worth having when the count is questioned later. In AMPthilly, scanning an asset’s QR label with a normal phone camera opens its profile in the browser, so a counter can confirm identity, owner, and status on the spot with no app install.
Common mistakes
- Counting from the desk - confirming items from memory or by asking around is asset verification theatre; if nobody stood in front of the item, it was not counted.
- Letting the record-keeper count alone - the person who maintains the register checking their own work is a classic gap in internal controls; pair them with someone independent.
- Fixing the register silently - adjust after investigating, never during the walk.
- Ignoring condition - a present-but-broken projector counted as fine just postpones the discovery.
- Counting while things move - deliveries, checkouts, and transfers mid-count create discrepancies that never existed.
Counts in practice
A count that happens reliably beats a perfect count that happens once. Put a recurring date in the calendar, keep scopes small enough to finish in a session, rotate counters so no one audits only their own records, and treat every discrepancy as a question about process - because a missing item is rarely the first thing that went wrong, just the first one anyone noticed.
Related terms
- Asset Reconciliation - resolving the differences a count uncovers
- Asset Verification - confirming individual assets exist as recorded
- Fixed Asset Audit - the formal review a good count prepares you for
- Audit Readiness - keeping records fit for inspection year-round
- Internal Controls - the checks and separations that keep counts honest