Category: Inventory Management

Inventory is usually the largest amount of money a retailer has sitting in one place — on shelves, in the stockroom and in boxes waiting to be unpacked. Retail inventory management is the set of habits and tools that make sure that money turns into sales instead of dust, write-offs or empty shelves.

This practical guide follows stock through the store: receiving, product identification, stock movements, valuation, replenishment, suppliers, audits and reporting. Each section ends with something you can apply this week, and the guide links to deeper articles on reorder points, ABC analysis, cycle counting and shrinkage.

The inventory cycle in one picture

Every product goes through the same loop: it is received, labelled and shelved, sold, counted and reordered. Good inventory management is simply making sure each step updates the same stock record accurately.

When the steps are connected — for example, when a POS system deducts stock at the till and adds it when deliveries are recorded — the stock figure on screen stays close to what is on the shelf. When steps happen on paper or in separate spreadsheets, the figures drift apart.

Receiving: where accuracy starts

Errors made at the back door spread through every report that follows. A disciplined receiving routine takes a few extra minutes per delivery and prevents hours of investigation later.

1. Compare against the order. Check the delivery note against your purchase order before signing.
2. Count physically. Count cases and, for high-value items, units. Do not rely on the supplier’s paperwork alone.
3. Inspect condition. Note damaged, expired or incorrect items on the delivery note and photograph them.
4. Record the purchase immediately. Enter quantities and purchase prices so stock and costs update the same day.
5. Label before shelving. Items without a scannable barcode get a label before they reach the sales floor.

Recording deliveries against a supplier keeps your cost prices current and your history useful for negotiations — see purchase management and supplier records.

SKU management and product identification

A SKU (stock-keeping unit) is your internal identifier for each distinct product you stock. Every variant that you need to count or price separately — size, colour, pack size — deserves its own SKU.

Use the manufacturer’s GTIN where it exists

Most branded products carry a GTIN (Global Trade Item Number) issued under GS1 standards and printed as a barcode — typically EAN-13 internationally or UPC-A (a 12-digit GTIN) in North America. GS1 guidance is clear that a changed product should get a new GTIN, and that outer cases usually carry a different GTIN from the consumer unit. Storing the GTIN on each SKU means a scan at the till or at receiving finds the right product instantly. For a full comparison of the identifiers, see SKU vs UPC vs EAN vs GTIN.

Label products that have no barcode

Loose, handmade or repackaged goods need internal barcodes. Generate them from the product record and print labels, so the code on the shelf always matches the system. See barcode labels and barcode scanning.

Good SKU practice	Why it matters
One SKU per sellable variant	Counts and reorders become precise
Consistent naming (Brand · Product · Size)	Faster search at the till
Category and brand on every SKU	Meaningful reports and ABC analysis
Cost and selling price stored	Margins and stock value are visible
No duplicate barcodes	Scans never pick the wrong product

Stock movements: every change has a reason

Stock should only change for a recorded reason. The common movement types are:

Movement	Direction	Typical source
Sale	−	POS transaction
Purchase / delivery	+	Supplier purchase
Customer return (resaleable)	+	Return at the till
Damage / expiry write-off	−	Manual adjustment with reason
Count correction	+ / −	Cycle count or stocktake
Transfer or internal use	−	Manual adjustment with reason

Reviewing the movement history of a product is the fastest way to explain a discrepancy. An activity log that records who made each manual adjustment discourages careless or dishonest changes — see staff permissions and activity log.

Stock valuation and key ratios

You need a stock value for your accounts and for decisions. The three common cost methods are:

• FIFO (first in, first out): the oldest purchase costs are assigned to sales first.
• Weighted average cost: each unit carries the average cost of all units on hand.
• Specific identification: each item’s actual cost is tracked — practical for high-value, serialised goods.

Weighted average example: you hold 100 units bought at $2.00 and receive 50 more at $2.30. The average cost is (100 × 2.00 + 50 × 2.30) ÷ 150 = $315 ÷ 150 = $2.10 per unit. Which method you must use for tax purposes depends on your country, so confirm with your accountant.

Inventory turnover and days of inventory

Inventory turnover = cost of goods sold ÷ average inventory value. Days of inventory = 365 ÷ turnover.

Example: a shop with $240,000 cost of goods sold in a year and an average inventory value of $40,000 has a turnover of 240,000 ÷ 40,000 = 6, or about 365 ÷ 6 ≈ 61 days of stock on hand. Compare the figure by category over time rather than against generic benchmarks.

Replenishment: deciding when and how much to order

The core question is simple — when should I reorder, and how much? The standard answer is the reorder point: reorder when stock falls to the expected demand during the supplier’s lead time plus a safety buffer.

Not every product deserves the same attention. ABC analysis ranks products by their annual value so you can apply tight reorder rules to the few items that matter most and simpler rules to the long tail. Low-stock lists in your POS are a practical daily trigger — see inventory management.

Working with suppliers

• Track actual lead times, not the promised ones. The reorder point depends on them.
• Record purchase prices on every delivery to spot price increases early.
• Note fill rate: how often deliveries arrive complete and correct.
• Agree on how damaged or short deliveries are credited, and keep evidence.
• Keep a second source for your most important (class A) products where possible.

Inventory audits: stocktakes and cycle counts

Even with perfect processes, records drift because of miscounts, unrecorded damage and theft. Two audit approaches keep them honest:

	Full stocktake	Cycle counting
What is counted	Everything at once	A small set of SKUs each day or week
Disruption	High — often requires closing	Low — done during normal operation
Error detection	Once or twice a year	Continuous
Best for	Year-end valuation	Day-to-day accuracy

Many retailers combine both: continuous cycle counts during the year and a full count at year-end. Our cycle counting guide shows how to schedule counts and measure accuracy, and the shrinkage prevention guide explains how to act on the losses you find.

Inventory reporting: the numbers to review

Report	Question it answers	Review
Low-stock list	What must I reorder now?	Daily
Sales by product / category	What is moving?	Weekly
Stock value by category	Where is my money sitting?	Monthly
Slow movers / no sales in 90 days	What should I discount or stop buying?	Monthly
Purchase history by supplier	What did I pay and when?	Before each order
Count adjustments	Where are records drifting?	Weekly

See which reports are available in each edition on the reports and analytics page.

Planning for seasons and promotions

Average demand hides peaks. Before a season or promotion, look at the same period last year, adjust for known changes (new products, price changes, local events), and order early enough to cover longer supplier lead times at busy periods. After the peak, review what sold through and what was left over; leftover seasonal stock is cheaper to clear early with a planned markdown than to store until next year.

Common inventory management mistakes

• Recording deliveries late. Stock looks low, so you reorder what is already in the back room.
• Correcting counts without investigating. The same error repeats next month.
• Managing every product the same way. Time spent on C items is time not spent on A items.
• Ignoring slow movers. Old stock ties up cash and shelf space that faster products could use.
• Relying on memory for supplier lead times. Reorder points built on guesses run out at the worst moment.

Retail inventory management checklist

• Every sellable variant has its own SKU, barcode, category and cost price.
• Deliveries are checked against the order and recorded the same day.
• Manual stock adjustments always carry a reason and a user.
• Low-stock items are reviewed daily; reorder points are set for key products.
• Products are classified A/B/C and reviewed at least twice a year.
• Cycle counts run every week; discrepancies are investigated, not just corrected.
• Slow movers are reviewed monthly.
• Backups of the inventory database run automatically.

Frequently asked questions

Related guides

• Reorder point formula
• ABC inventory analysis
• Inventory cycle counting
• Retail shrinkage prevention
• What is a POS system?

Sources and further reading

• GS1 UK — FAQ on GTINs and barcodes
• GS1 New Zealand — Bar coding basic user guide
• Zoho Inventory — ABC analysis in inventory management
• Cleverence — Safety stock and reorder point calculation

Order too late and customers find an empty shelf. Order too early and cash sits in the stockroom. The reorder point tells you the stock level at which to place the next order so that it arrives just before you run out — with a small buffer for the unexpected.

This guide explains the reorder point formula, how to measure demand and supplier lead time, three ways to calculate safety stock, and four fully worked examples. All calculations have been checked; you can reproduce them in a spreadsheet.

The reorder point formula

• Average daily demand (d): how many units you sell per day on average.
• Lead time (L): the number of days between placing an order and the stock being available to sell.
• Safety stock (SS): extra units held to absorb higher-than-average demand or late deliveries.

The first part, d × L, is the stock you expect to sell while waiting for the delivery. Safety stock covers the uncertainty around that expectation.

Measuring demand correctly

Use your own sales history rather than guesses. A POS that records every sale by product makes this straightforward.

• Use a representative period: the last 8–12 weeks for steady products; the same season last year for seasonal products.
• Use units, not revenue. Price changes distort revenue-based estimates.
• Exclude one-off events such as a single bulk order from a business customer, or treat them separately.
• Count days the shop is open. If you trade six days a week, divide weekly sales by six, and use lead time in trading days.
• Watch for stockouts in the history: days with zero stock show zero sales, which understates real demand.

Sales reports by product and period make this quick — see reports and analytics.

Measuring supplier lead time

Lead time is not only the shipping time. It runs from the moment you decide to order until the stock is on the shelf:

Component	Example
Time to place the order	1 day (you order on your weekly ordering day)
Supplier processing	2 days
Transport	3 days
Receiving, checking and shelving	1 day
Total lead time	7 days

Record the actual date you ordered and the date stock was available for each delivery. Recording supplier purchases with dates gives you this history automatically — see purchase management and supplier records.

Safety stock: three methods

1. Fixed buffer (simple)

Choose a number of days of cover — for example, three days of average demand. Easy to apply and explain, but it ignores how variable demand actually is.

2. Statistical method (demand varies)

The service level here is the probability of not running out during a replenishment cycle. Common Z values from the standard normal distribution:

Cycle service level	Z value
90%	1.28
95%	1.645
97.5%	1.96
99%	2.33

Use the same time unit throughout. If σ is measured per day, lead time must be in days; mixing a weekly or monthly standard deviation with a daily lead time is a common error that inflates safety stock dramatically.

3. Statistical method (demand and lead time vary)

Use this version when suppliers are unreliable: the d² × σ_L² term often dominates, which is why late deliveries hurt more than demand noise.

Alternative: the max–average method

Without enough data for statistics, some retailers use SS = (maximum daily demand × maximum lead time) − (average daily demand × average lead time). It is simple but conservative and tends to overstate stock.

Worked examples

Example 1 — Fixed buffer

A café sells an average of 12 bags of coffee beans per day. The roaster delivers in 5 days. The owner keeps 20 bags as safety stock.

ROP = 12 × 5 + 20 = 60 + 20 = 80 bags. When stock falls to 80, place the order.

Example 2 — Statistical safety stock

A grocery store sells an average of 15 units/day of a cooking oil, with a standard deviation of 4 units/day. Lead time is 9 days. Target service level: 95% (Z = 1.645).

1. Safety stock = 1.645 × 4 × √9 = 1.645 × 4 × 3 = 19.74 → round up to 20 units.
2. Lead-time demand = 15 × 9 = 135 units.
3. ROP = 135 + 20 = 155 units.

Example 3 — Demand and lead time both vary

A hardware store sells 20 units/day of a fastener pack (σ_d = 5). The supplier’s lead time averages 7 days but varies (σ_L = 2 days). Target: 95% (Z = 1.645).

1. Combined deviation = √(7 × 5² + 20² × 2²) = √(175 + 1,600) = √1,775 ≈ 42.13.
2. Safety stock = 1.645 × 42.13 ≈ 69.3 → 70 units.
3. ROP = 20 × 7 + 70 = 140 + 70 = 210 units.

Notice that the lead-time variability (1,600) contributes far more than demand variability (175). Improving supplier reliability would reduce this buffer more than better forecasting.

Example 4 — Max–average method

Maximum daily demand 30, maximum lead time 10 days; average demand 20, average lead time 7 days. Safety stock = 30 × 10 − 20 × 7 = 300 − 140 = 160 units — much higher than Example 3, illustrating why this method is conservative.

Calculating reorder points in a spreadsheet

You do not need special software to start. Export daily unit sales for each product, then build one row per SKU:

Column	Content	Example formula (Excel / Google Sheets)
A	Daily sales history (one column per day)	—
Avg demand (d)	Average units per trading day	=AVERAGE(B2:BI2)
σd	Standard deviation of daily demand	=STDEV.S(B2:BI2)
Lead time (L)	Days from order to shelf	entered manually
Z	Service level factor	=NORM.S.INV(0.95) → 1.645
Safety stock	Z × σd × √L, rounded up	=ROUNDUP(Z*σd*SQRT(L),0)
Reorder point	d × L + safety stock	=ROUNDUP(d*L,0)+SS

NORM.S.INV returns the Z value for any service level, so you can test what moving from 95% to 97.5% does to stock — in Example 2 it would raise safety stock from about 20 to about 24 units (1.96 × 4 × 3 = 23.5).

Special cases

• Seasonal products: calculate demand from the same season last year, not from the last few weeks.
• New products: start with a conservative estimate based on similar items and review after four to six weeks of sales.
• Promotions: add the expected promotional uplift to demand during the promotion period only.
• Perishable goods: safety stock is limited by shelf life; a lower service level is often the right trade-off.
• Slow movers (a few units a month): statistical formulas are unreliable; use a simple minimum such as one or two units on hand.
• Weekly ordering days: if you can only order once a week, add the review period to the lead time (periodic review), or you will run out between order days.

How much to order once you hit the reorder point

The reorder point tells you when; you still need to decide how much. Common retail approaches:

• Fixed quantity: always order the same amount (for example, a full case or a supplier minimum).
• Order-up-to level: order enough to bring stock back to a target maximum, such as two weeks of demand plus safety stock.
• Days of cover: order a number of days of expected demand, adjusted for upcoming promotions or seasons.

Balance three costs: the cost of placing orders, the cost of holding stock, and the cost of running out. Case sizes and minimum order values usually decide the final number.

Reorder point vs. min/max

Many POS and inventory tools use a minimum and maximum per product. The two ideas fit together: the minimum is effectively your reorder point, and the maximum is your order-up-to level. When stock falls to the minimum, you order enough to reach the maximum. Using the method in this guide to set the minimum — instead of a round number picked once and never revisited — is what turns a min/max setting into a real replenishment policy.

Revisit both values when demand, lead time or case sizes change, and after every peak season.

Preventing stockouts in practice

• Set reorder points for your class A products first — see ABC inventory analysis.
• Review the low-stock list every day at the same time.
• Recalculate demand and lead time every quarter, and before peak seasons.
• Keep stock records accurate with regular cycle counts — a reorder point is useless if the stock figure is wrong.
• Record deliveries on the day they arrive so the system stock is current.
• Agree on emergency options with key suppliers.

For the bigger picture, read the retail inventory management guide. Low-stock visibility is part of inventory management in the POS.

Frequently asked questions

Related guides

• Retail inventory management guide
• ABC inventory analysis
• Inventory cycle counting
• How to choose a POS system

Sources and further reading

• Cleverence — How to calculate safety stock and reorder point
• Cleverence — Safety stock and reorder point calculation guide
• Ecosire — EOQ, safety stock and reorder point guide
• NIST/SEMATECH e-Handbook — Normal distribution

In most shops, a small number of products account for a large share of the money flowing through inventory. ABC inventory analysis makes that visible, so you can spend your limited time controlling the products that matter most — and stop over-managing the ones that do not.

This guide explains the ABC method step by step, shows a complete worked example with ten products, and turns the result into concrete policies for ordering, counting and pricing.

What is ABC inventory analysis?

ABC analysis ranks products by their annual consumption value — how much money moves through each product in a year — and groups them into three classes:

Class	Typical share of SKUs	Typical share of value	Management attention
A	about 10–20%	about 70–80%	High — tight control
B	about 20–30%	about 15–25%	Medium
C	about 50–70%	about 5–10%	Low — simple rules

The idea follows the Pareto principle (“80/20 rule”). The percentages are conventions, not laws: practitioners use slightly different thresholds, and your own data decides where the natural breaks are.

Annual consumption value

Use cost rather than selling price when the goal is inventory investment and control. Some retailers rank by gross profit or revenue instead when the goal is commercial focus; the method is the same, but be consistent and state which basis you used.

You need two data points per SKU: the unit cost (from your purchase records) and the number of units sold in the last twelve months (from your sales reports). See purchase management and reports.

How to do an ABC analysis in five steps

1. Export each SKU’s unit cost and units sold over the last 12 months.
2. Calculate ACV for each SKU: unit cost × annual units.
3. Sort SKUs from highest to lowest ACV.
4. Calculate each SKU’s share of total ACV and the cumulative percentage down the list.
5. Classify: SKUs up to roughly 70–80% cumulative value are A, the next band up to roughly 90–95% are B, and the rest are C. Adjust the cut-offs to the natural breaks in your data.

A spreadsheet is enough. Sort by ACV, add a running total column and divide by the grand total.

Worked example: a coffee equipment shop

A small specialty coffee shop sells beans, equipment and accessories. Here are its ten SKUs, ranked by annual consumption value (unit cost × annual units). All figures are calculated and rounded to one decimal place.

Rank	SKU	Unit cost	Annual units	ACV	Share	Cumulative	Class
1	Premium beans 1 kg	$14.00	2,600	$36,400	32.2%	32.2%	A
2	Espresso machine	$420.00	60	$25,200	22.3%	54.6%	A
3	Coffee grinder	$95.00	180	$17,100	15.2%	69.7%	A
4	Milk frother	$38.00	320	$12,160	10.8%	80.5%	B
5	Paper cups (box of 1,000)	$22.00	410	$9,020	8.0%	88.5%	B
6	Ceramic mug	$6.50	900	$5,850	5.2%	93.7%	B
7	Descaler	$5.20	600	$3,120	2.8%	96.4%	C
8	Filter papers	$1.80	1,500	$2,700	2.4%	98.8%	C
9	Cleaning brush	$2.40	350	$840	0.7%	99.6%	C
10	Gift card sleeves	$0.60	800	$480	0.4%	100.0%	C

Total annual consumption value: $112,870.

Reading the result

• Class A — 3 SKUs (30% of items) carry 69.7% of value. Two of them — the espresso machine (60 units) and the grinder (180 units) — are the two least frequently sold items in the whole range, yet they rank second and third by value.
• Class B — 3 SKUs (30% of items) carry 24.0% of value (93.7% − 69.7%).
• Class C — 4 SKUs (40% of items) carry just 6.3% of value. Filter papers sell 1,500 units a year but matter little financially.

With only ten SKUs the split is less extreme than in a real shop with hundreds or thousands of products, where the long tail of C items is usually much larger.

Turning classes into management policies

Policy	Class A	Class B	Class C
Reorder method	Calculated reorder point with safety stock	Reorder point, reviewed quarterly	Simple min/max or periodic top-up
Service level target	Highest (e.g. 97–99%)	Medium (e.g. 95%)	Lower (e.g. 90%)
Cycle count frequency	Monthly or more	Quarterly	Once or twice a year
Count tolerance	Exact match	Small tolerance	Wider tolerance
Supplier management	Track lead time; second source	Regular review	Consolidate orders
Shelf position / security	Prime, visible, protected	Standard	Standard
Review frequency	Weekly	Monthly	Quarterly

The service-level and frequency figures are starting points, not standards. Adjust them to your margins, space and staffing. For counting schedules, see inventory cycle counting.

ABC analysis in a spreadsheet

Column	Formula (Excel / Google Sheets)
C: Annual consumption value	=A2*B2 (unit cost × annual units)
Sort	Sort the whole table by column C, largest to smallest
D: Share of total	=C2/SUM($C$2:$C$1001)
E: Cumulative share	=SUM($D$2:D2)
F: Class	=IF(E2<=0.7,”A”,IF(E2<=0.95,”B”,”C”))

Change 0.7 and 0.95 to the cut-offs you choose. In the worked example above, these thresholds give three A items, three B items and four C items.

Using ABC to manage cash and purchasing

ABC is not only a counting tool. Because A items hold most of the money moving through inventory, small improvements on them have the biggest effect on cash flow:

• Order A items more often in smaller quantities where suppliers allow, so less cash sits in stock.
• Negotiate on A items: a few percent on a high-value line is worth more than large discounts on C items.
• Order C items less often in larger quantities to reduce ordering effort; the cash impact is small.
• Review slow-moving B and C items for discontinuation when space is tight.

Illustrative scenario

The following is a hypothetical illustration, not a customer case. A hardware store with 4,000 SKUs runs its first ABC analysis and finds that roughly 400 items make up most of its annual consumption value. It moves those items to weekly reorder reviews and monthly counts, sets reorder points for them, and switches thousands of low-value fasteners and accessories to a simple monthly top-up. The owner spends less total time on stock — but now spends it where the money is.

Refinements: XYZ, criticality and new products

• XYZ analysis adds demand variability: X items sell steadily, Z items sell erratically. An “AZ” item (valuable and unpredictable) needs more safety stock than an “AX” item.
• Criticality overrides: a low-value C item can still be essential — the paper cups a café cannot trade without. Flag such items manually.
• New products lack a year of history. Classify them provisionally on expected sales and revisit after three months.
• Seasonality: a seasonal product may be A in December and C in March. Run the analysis on the relevant period.

What ABC changes in a store’s daily routine

Classification only pays off when it changes what people do. In practice, a retailer using ABC typically ends up with three different rhythms:

Routine	Class A	Class B	Class C
Low-stock check	Every day	Twice a week	Weekly
Ordering	Per reorder point, possibly several times a week	Weekly order	Monthly top-up
Counting	Rolling monthly counts	Quarterly	Once or twice a year
Price checks vs. suppliers	Every delivery	Quarterly	Yearly

These rhythms are examples to adapt, not rules. The point is that the shop owner’s limited attention follows the money. Keep the class visible — for example as a prefix in an internal note or category tag — so staff know which products deserve extra care when stock is low or a delivery is short.

Where the data comes from

Everything ABC needs is already produced by a POS that records sales by product and purchases with cost prices: units sold per product over twelve months, and the latest unit cost. Export both, run the spreadsheet steps above, and repeat twice a year. Accurate inputs depend on accurate stock and cost records — another reason to keep up regular cycle counts and to record every delivery.

Common ABC analysis mistakes

• Ranking by units instead of value. High-volume, low-cost items look important when they are not.
• Never updating the classes. Review at least twice a year; product ranges and prices change.
• Ignoring gross margin entirely. For buying decisions, also look at profit contribution, not just cost.
• Using inaccurate stock or cost data. The analysis is only as good as your records — keep them clean with regular counts and recorded purchase prices.
• Treating C items as unimportant to customers. Missing small items still frustrates shoppers; keep simple, reliable rules for them.

ABC analysis checklist

• Export 12 months of units sold and current unit cost for every SKU.
• Calculate ACV, sort, compute cumulative percentages.
• Set cut-offs at natural breaks (around 70–80% and 90–95%).
• Apply different reorder, count and review policies by class.
• Flag critical C items manually.
• Schedule the next review date (every six months, or before peak season).

For the overall framework, see the retail inventory management guide. Product categories, brands and cost prices are managed in inventory management.

Frequently asked questions

Related guides

• Reorder point formula
• Inventory cycle counting
• Retail inventory management guide
• Retail shrinkage prevention

Sources and further reading

• Zoho Inventory — ABC analysis in inventory management
• DCL Logistics — ABC analysis: meaning, advantages and challenges
• Cleverence — What is ABC analysis of inventory?

A stock figure is only useful if you can trust it. Reorder points, low-stock alerts and stock valuation all assume that the number on screen matches the number on the shelf. Inventory cycle counting keeps those numbers honest by counting a small part of your stock continuously, instead of everything once a year.

This guide compares cycle counting with a full physical inventory, shows how to build a counting schedule, how to handle discrepancies, how to measure accuracy, and gives a step-by-step procedure your team can follow.

Physical inventory vs. cycle counting

	Full physical inventory	Cycle counting
Scope	All stock at once	A subset of SKUs on a schedule
Frequency	Usually once or twice a year	Daily or weekly
Disruption	High — store often closes or slows down	Low — fits into normal operations
Error detection	Late; causes are hard to trace	Early; causes are still fresh
Staff fatigue	High, which itself causes counting errors	Low — short, focused counts
Main purpose	Year-end valuation, accounting requirement	Ongoing record accuracy

Cycle counting does not always replace the annual count — your accountant or local rules may require one — but it makes the annual count faster and less surprising, because records are already accurate.

Ways to select what to count

• ABC-based counting: count high-value A items often, C items rarely. The most common retail approach — see ABC inventory analysis.
• Location-based: count one aisle, shelf or bay at a time until the whole store is covered.
• Opportunity-based: count an item when stock is low (fewer units to count), when it reaches zero, or when a delivery arrives.
• Exception-based: count items flagged by the system — negative stock, unusual adjustments, sales with zero recorded stock.
• Random sampling: a few random SKUs each day to measure overall accuracy without bias.

Most stores combine ABC-based scheduling with exception counts.

Building a cycle count schedule

Decide how often each class is counted, then calculate the daily workload.

Example: a store has 1,000 SKUs: 100 A items, 300 B items and 600 C items. It counts A items monthly (12 times a year), B items quarterly (4 times) and C items twice a year.

Class	SKUs	Counts per year each	Total counts per year
A	100	12	1,200
B	300	4	1,200
C	600	2	1,200
Total	1,000		3,600

With about 250 counting days a year, that is 3,600 ÷ 250 = 14.4 SKUs per day — roughly 15 counts, which a trained employee can usually complete in a short session before opening. Adjust frequencies until the daily load fits your staffing.

Step-by-step cycle count procedure

1. Generate the count list for today from your schedule.
2. Count blind: the counter records what they find without seeing the system quantity, which avoids “confirming” a wrong number.
3. Count every location where the SKU can be — shelf, back stock, display, returns area.
4. Compare the count with the system quantity at the same moment.
5. Recount if the difference exceeds the tolerance — ideally by a different person.
6. Investigate confirmed differences before adjusting (see below).
7. Adjust the record with a reason code — count correction, damage, theft suspected, receiving error — and the name of the person approving it.
8. Record the result for your accuracy metrics.

Using the barcode scanner to identify each product during the count reduces mix-ups between similar items — see barcode scanning.

Investigating and reconciling discrepancies

A discrepancy is a symptom. Correcting the number without finding the cause means it will happen again. Check these in order:

Check	What you might find
Recent deliveries	Delivery received but not recorded, or recorded twice
Similar products	Wrong variant scanned at the till (size, flavour, colour)
Other locations	Stock in the back room, on a display or in returns
Recent returns	Returned item restocked but not recorded, or recorded but not restocked
Manual adjustments	Unexplained adjustments by a user
Damage and expiry	Products thrown away without a write-off
Theft	Pattern of losses on easily concealed, high-value items

Movement history and an activity log of who changed what make this investigation fast — see inventory management and activity log. Repeated unexplained losses are a shrinkage signal; see retail shrinkage prevention.

Measuring inventory accuracy

Example: this month you counted 200 SKUs; 186 matched the system within tolerance. Accuracy = 186 ÷ 200 × 100 = 93%.

Setting tolerances

Practitioners commonly apply tighter tolerances to higher-value items — often an exact match for A items and a small percentage or unit tolerance for low-value C items. Some organisations remove tolerances entirely for their accuracy metric. Decide your rules, write them down and apply them consistently, otherwise the metric is meaningless.

Other useful measures

• Value of adjustments per month (gains and losses separately, not netted).
• Coverage: share of scheduled counts actually completed.
• Repeat offenders: SKUs with discrepancies in consecutive counts.
• Accuracy by class and by category to see where processes break down.

Preparing for a count

• Freeze movements for the counted SKUs: finish receiving and put away deliveries before counting those items.
• Agree the cut-off: sales after the count starts must be accounted for, or count before opening.
• Tidy first: group identical products together and return misplaced items to their home.
• Prepare tools: printed or on-screen count list, scanner, and a clear rule for units vs. packs.
• Know where stock hides: back stock, window displays, the returns shelf, the repair or damaged area.

Who should count?

Separation of duties matters. Ideally, the person who receives deliveries or manages a category should not be the only person counting it, and the person approving adjustments should not be the one who counted. In a very small shop this may not be possible every day, but alternating counters and having the owner spot-check high-value items achieves a similar effect.

Role	Responsibility
Counter	Counts blind and records quantities
Recounter	Recounts items outside tolerance
Approver	Reviews the cause and approves the adjustment
Owner / manager	Reviews accuracy metrics monthly

Common cycle counting mistakes

• Counting while selling the same items — creates false variances.
• Showing the expected quantity to the counter — people tend to confirm it.
• Adjusting without a reason code — destroys the information you need to prevent the next error.
• Netting gains and losses — a +5 on one SKU and −5 on another is two errors, not zero.
• Skipping counts when busy — coverage drops, and so does trust in the numbers.

Improving accuracy over time

• Fix receiving first: most discrepancies start at the back door. Record deliveries on arrival — see purchases.
• Scan, don’t key: scanning every item at the till eliminates “same price, wrong product” errors.
• Label everything: products without barcodes get internal labels — see barcode labels.
• One home per product: fewer storage locations means fewer missed units.
• Train counters on units of measure: single items vs. packs vs. cases.
• Review the accuracy trend monthly and celebrate improvement — counts done well are boring.

Cycle counts and the year-end count

If your accountant still requires a full count at year-end, cycle counting makes it easier rather than redundant. Because records are already close to reality, the year-end count becomes a confirmation exercise: fewer recounts, smaller adjustments and fewer surprises in the stock valuation. Some businesses, with their accountant’s agreement, use documented cycle count results to support the year-end figure; whether that is acceptable depends on your local rules, so ask before relying on it.

Cycle counting checklist

• Classify SKUs A/B/C and set a counting frequency for each class.
• Calculate the daily count load and assign a responsible person.
• Count blind, at a quiet time, in every location.
• Recount differences beyond tolerance before adjusting.
• Investigate the cause; adjust with a reason code and approver.
• Track accuracy %, adjustment value and coverage every month.

For the complete framework, read the retail inventory management guide, and keep reorder points reliable by keeping counts accurate.

Frequently asked questions

Related guides

• ABC inventory analysis
• Retail shrinkage prevention
• Reorder point formula
• POS hardware setup checklist

Sources and further reading

• Cleverence — Cycle counting for accuracy, methods and KPIs
• Oracle Inventory documentation — Cycle count approvals and tolerances

Shrinkage — usually shortened to “shrink” — is inventory you paid for but can no longer sell or account for. It includes theft, but also receiving mistakes, damage, expired goods, return abuse and simple record-keeping errors. Because shrink is invisible until you count, many small retailers underestimate it.

This guide explains how to measure shrink, where it comes from, and the practical controls — from receiving checks to audit trails — that reduce it without turning your shop into a fortress.

What is retail shrinkage?

Shrink is the difference between the inventory your records say you should have and the inventory you actually have, valued at cost and usually expressed as a percentage of sales for the period.

Example: your records show $52,000 of stock at cost. A full count finds $51,220. Shrink = $52,000 − $51,220 = $780. Net sales for the period were $48,750, so the shrink rate is 780 ÷ 48,750 × 100 = 1.6%.

For context, the U.S. National Retail Federation’s 2023 National Retail Security Survey reported an average shrink rate of 1.6% of sales for fiscal year 2022, up from 1.4% the previous year. Industry averages vary widely by sector and store, so track your own trend rather than chasing a benchmark.

The main causes of inventory loss

Theft is a large part of shrink, but not all of it. In the same NRF survey, external and internal theft together accounted for about two-thirds of shrink; the rest came from causes such as process and administrative errors, damage and vendor fraud. That split matters: process problems are usually the cheapest to fix.

Cause	Examples	Typical signal
External theft	Shoplifting, organised theft	Losses on small, high-value, easily resold items
Internal theft	Staff taking stock or cash, “sweethearting” (undercharging friends)	Unusual voids, discounts or refunds by one user
Receiving errors	Short deliveries signed as complete, wrong items	Losses that appear right after deliveries
Administrative errors	Wrong product scanned, unrecorded transfers, pricing mistakes	Gains on one variant, losses on a similar one
Damage and expiry	Breakage, spoiled food, products thrown away without a write-off	Losses in fragile or dated categories
Return abuse	Returns without receipt, used items returned as new	High returns on specific products or staff shifts
Vendor fraud	Deliberate short shipments or invoice errors	Repeated discrepancies with one supplier

Reducing theft on the sales floor

• Design for visibility: keep sightlines clear from the counter, avoid tall displays near the door, and place high-value items near staff.
• Greet customers: attentive service is one of the simplest deterrents.
• Protect high-risk items: locked cases, display-only packaging or keeping stock behind the counter for small, expensive products.
• Use your ABC data: focus security on class A items that are also easy to conceal — see ABC inventory analysis.
• Count high-risk items frequently so losses are spotted within days, not months.

Receiving controls

Every unit lost at the back door is recorded as stock you never had. A short receiving routine prevents it:

• Check delivery notes against the purchase order before signing.
• Count cases — and units for high-value goods — in front of the driver when possible.
• Note damaged or missing items on the delivery note; photograph them.
• Record the purchase in the system the same day with actual quantities received.
• Keep delivery areas closed to customers and visitors.
• Rotate receiving duties, and spot-check deliveries received by others.

Recording purchases against suppliers also shows which suppliers deliver short repeatedly — see purchase management and supplier records.

Damaged and expired products

• Record every write-off with a reason (damaged, expired, used as tester). Unrecorded disposal looks exactly like theft in the numbers.
• Rotate stock first-in, first-out for dated products.
• Train handling for fragile goods and review storage that causes breakage.
• Claim from suppliers for goods damaged in transit, with evidence.

Returns and refunds

• Require a receipt or a findable sale for refunds; reprinting receipts from sales history makes genuine returns easy.
• Inspect before restocking: decide on the spot whether an item is resaleable or a write-off, and record which.
• Restrict refunds by role: only trusted users can approve refunds above a set value.
• Review refunds per user weekly; patterns stand out quickly.

Internal controls and audit trails

Most internal loss is opportunistic. Controls that make actions traceable remove the opportunity without accusing anyone:

Control	What it prevents
Individual logins for every employee	Anonymous voids, discounts and refunds
Role-based permissions	Cashiers changing prices or deleting sales
Activity log of sensitive actions	Untraceable manual stock adjustments
Manager approval for large discounts and refunds	Sweethearting and refund fraud
Cash counted per drawer and per shift	Unexplained cash differences
Separation of duties	One person receiving, adjusting and counting the same stock

An activity log that records which user performed each important action is the backbone of an audit trail — see staff, permissions and activity log.

Cash handling at the till

• One drawer, one responsible person per shift where possible.
• Count the float at the start and the drawer at the end of every shift, and compare with recorded cash sales.
• Record every payment method correctly so card and cash totals reconcile.
• Limit “no sale” drawer openings and review them.
• Remove excess cash from the drawer during busy days.

Reconcile cash every day

Cash differences are a form of loss that daily routines catch quickly. A consistent end-of-day count, with differences recorded by cashier, is explained step by step in our cash register reconciliation guide.

Training and culture

Controls work best when staff understand why they exist. Explain that individual logins and adjustment reasons protect honest employees as much as the business: when every action is traceable, nobody is suspected without evidence. Train new staff on receiving, write-offs and refunds in their first week, and share shrink results with the team — improvements are easier to sustain when people can see them.

Illustrative scenario

The following is a hypothetical illustration, not a customer case. A small cosmetics shop notices repeated losses on a few premium skincare lines. Weekly counts of those items show that losses cluster on delivery days. Investigation finds that cartons were signed for without being opened, and some arrived short. Introducing a receiving check — and moving testers and high-value items closer to the counter — addresses both process loss and opportunistic theft.

Common shrink-prevention mistakes

• Assuming all shrink is theft — and ignoring cheaper process fixes.
• Counting only once a year — losses are discovered months after they happen.
• Shared logins — make every investigation guesswork.
• Throwing damaged goods away without recording them — inflates “unknown” loss.
• Security measures that block service — locking everything away can cost more in lost sales than it saves.

Detecting shrink early

The earlier you see a loss, the easier it is to explain. Combine regular counting with a few reports:

• Cycle counts on high-risk and high-value items — see inventory cycle counting.
• Adjustment report by reason, product and user.
• Voids, discounts and refunds by user and shift.
• Negative or impossible stock (sales recorded when stock was zero) — often a scanning or receiving error.
• Category shrink trend per month.

See which reports are included on the reports and analytics page.

Measuring shrink by category

A single store-wide shrink rate hides where the problem is. Calculate shrink separately for your main categories — or at least for your class A products — using the same formula: recorded value minus counted value, divided by the category’s sales for the period. A category with a much higher rate than the rest of the store points you to a specific process (receiving, storage, display or returns) to investigate first. Track each category month by month so you can see whether the controls you introduce are working.

30-day shrink reduction plan

Week	Actions
Week 1	Measure: full count of your top 50 high-value SKUs; calculate shrink for each; set up individual logins.
Week 2	Receiving: introduce the receiving checklist; record every delivery on arrival.
Week 3	Controls: restrict refunds and large discounts by role; start weekly reviews of voids, refunds and adjustments.
Week 4	Floor: move high-risk items to visible or protected spots; recount the top 50 SKUs and compare with week 1.

Keep the cycle going: monthly shrink by category, quarterly review of controls, and a full count at year-end. For the overall framework, see the retail inventory management guide.

Frequently asked questions

Related guides

• Inventory cycle counting
• ABC inventory analysis
• Retail inventory management guide
• How to choose a POS system

Sources and further reading

• National Retail Federation — National Retail Security Survey 2023
• NRF press release — Retail shrink nearly a $100B problem
• GS1 UK — FAQ on product identification