The Inventory Count SOP

Purpose

Define how physical stock gets counted and reconciled to the system. A standard operating procedure (SOP) is the written, repeatable process everyone follows the same way, so the count is trustworthy regardless of who did it. A stock system that is never counted against reality drifts until it is useless: it says a part is on hand that is not, so a job stalls, or it hides shrinkage that has been walking out the back for months. A disciplined count catches the drift, quantifies the loss, and keeps the numbers the ordering process depends on honest.

Scope

Applies to counting physical inventory wherever it lives:

• Warehouse and shop shelf stock
• Truck and van stock
• Staged job materials held but not yet installed

Covers both the cycle count (a rolling subset counted often) and the full count (everything, less often). Excludes fixed assets and tools, which are tracked by assignment, not counted as stock.

Roles and responsibilities

Procedure

Step 1: Schedule the count and pick the type

Set a fixed cadence so counting is routine, not a fire drill:

• Cycle count: a rolling slice of stock counted on a regular short interval, weighted so fast-moving and high-value items come around more often.
• Full count: everything, on a longer interval, typically tied to the accounting period.

A cycle count catches most drift with far less disruption than shutting down for a full count.

Step 2: Freeze movement in the count area

Nothing moves in or out of the area being counted while the count is live. A part pulled for a job mid-count double-counts or vanishes and creates a phantom variance. Either count outside working hours or fence off the area and pause pulls from it.

Step 3: Count blind, then record

The counter counts what is physically there without looking at the system quantity first. Counting blind removes the pull to "confirm" the expected number instead of counting honestly. Record the raw count against each item and location.

Step 4: Compare to the system

Match the physical count to the system quantity. Items that agree are done. Items that differ go on a variance list. Small, expected variances on high-turn consumables are normal. A large gap, or any gap on a high-value item, gets attention.

Step 5: Recount every variance before adjusting

A different person recounts each variance. Most variances are miscounts, a mislabeled bin, or stock in the wrong location, not real loss. Never adjust the system off a single count. The recount confirms the number before anyone touches the record.

Step 6: Adjust and code the reason

Once confirmed, the manager approves the adjustment and codes why: shrinkage, damage, an uncounted job usage, a receiving error, or a return not processed. An adjustment with no reason code hides the pattern that would tell you where stock is actually leaking.

Step 7: Report shrinkage and act on the pattern

Total the confirmed shrinkage and look for the pattern. Loss concentrated on one item, one truck, or one location points at a cause: a broken process, a labeling problem, or theft. The count is only worth the effort if the pattern it reveals drives a fix.

Records and retention

Keep count sheets, variance lists, and approved adjustments with their reason codes. The history shows shrinkage trending over time and is the backup for the inventory value on the books.

Common count failures

• Movement not frozen. Parts counted while jobs pull from the same bins produce phantom variances.
• Counting with the system number in hand. The count confirms the expectation instead of reality.
• No recount. The system gets adjusted off a miscount, making the record worse.
• No reason codes. Shrinkage is recorded but never explained, so it never gets fixed.

References

• SBA guidance on inventory management and controls for small businesses
• IRS Publication 538, Accounting Periods and Methods, for inventory accounting
• See related: The Parts Ordering SOP
• Trade-standard practice for cycle counting and physical inventory