How often should a small business count inventory?

Cycle count continuously: top-selling items monthly, mid-tier quarterly, slow movers twice a year. That removes the need for an annual shutdown.

What is a normal inventory shrinkage rate?

Most small retailers land between 1% and 3% of cost of goods sold. Above 3% usually points to a process problem rather than theft.

Should counters see the expected quantity?

No. Count blind. Given an expected number, people tend to find it, which hides the variance you are trying to measure.

Small Business Inventory Audit: A Practical Guide

The Small Business Inventory Audit: A Practical Guide

An inventory audit answers one question: does what the system says match what is on the shelf? For most small businesses the answer is no, and the gap is usually 2-8% of stock value. Finding that gap is often the fastest profit improvement available to you.

Why the numbers drift

Stock records drift for boring reasons, not dramatic ones:

Items sold but not scanned during a rush
Damaged goods binned without a write-off
Supplier shorted the delivery and nobody checked
Returns put back on the shelf but never re-added
Samples, staff purchases, and display units untracked
Genuine theft, usually the smallest slice for a small shop

Cycle counts beat annual counts

A full annual count means closing the shop, exhausting everyone, and getting sloppy data by hour six. Cycle counting spreads the same work across the year: count a small slice of stock every week, and every item gets counted several times a year.

A simple ABC rotation works well:

A items (top 20% of items, ~80% of sales value) — count monthly
B items (next 30%) — count quarterly
C items (remaining 50%) — count twice a year

Twenty minutes before opening, three or four times a week, covers a surprisingly large catalogue.

How to run a clean count

Freeze movement. Count before opening or after closing. Never count a section while it is being sold from.
Count blind. Do not show the counter the expected quantity. Given a number, people find that number.
Scan, do not type. A barcode scan removes the two biggest error sources — misreading SKUs and typos.
Count twice on variances. Any discrepancy above your threshold gets a second count by a different person before adjusting.
Record the reason. "Adjust -4, damaged in transit" is data. A silent adjustment is not.

Setting a variance threshold

Small shops usually set a tolerance of about 2% by unit count or a fixed dollar figure per SKU, whichever is tighter. Anything inside tolerance gets adjusted quietly; anything outside gets investigated. Without a threshold you either investigate everything (nobody sustains this) or nothing (the point is lost).

Turning findings into money

An audit that ends in an adjusted number is half-finished. Look at the pattern:

Shrinkage concentrated in small, high-value items — move them behind the counter or into locked display.
Shrinkage concentrated at one time of day — that is a process problem at the busiest shift, usually unscanned sales.
Consistent supplier shortages — start counting deliveries at receipt and claim credits.
Damaged stock in one category — fix storage or packaging, not the count.

Valuation basics

Your inventory value feeds directly into your tax return and your true margin. Pick one costing method and stay with it:

FIFO (first in, first out) — assumes oldest stock sells first; the most common and usually the most intuitive.
Weighted average — smooths out price swings; easy to run in software.

Whatever you choose, apply it consistently. Switching methods mid-year makes year-over-year comparison meaningless and can raise questions with your accountant.

Tooling that makes this painless

Counting is only hard when the record-keeping is hard. With ExpressBizTools you scan an item with your phone camera or a USB scanner, enter the counted quantity, and the stock level updates with a timestamp. Low-stock alerts then tell you what to reorder before you run out, and the reports view shows movement over time.

Pair it with two cheap physical upgrades: printed shelf labels with scannable barcodes, and labelled bins for the backroom. Most counting time is spent looking for things, not counting them.

A repeatable monthly rhythm

Week 1 — count A items
Week 2 — count one B category
Week 3 — count one B category and reconcile supplier deliveries
Week 4 — count a C sample, review adjustments, write down dead stock

Ninety minutes a month is enough for most shops under 1,000 SKUs. The return — accurate reorder points, fewer stockouts, less capital sitting in dead stock — is far bigger than the time cost.