Merchandise, Inventory and On-Shelf Availability

Introduction

Retail profitability depends not only on selling the right products, but also on making them available at the right place and time. Merchandise decisions determine what the retailer offers; inventory decisions determine how much capital is committed to those products; replenishment decisions determine whether customers find the products when they intend to buy them. These decisions are interdependent. Excess stock creates holding costs, markdowns, obsolescence, waste and shrinkage, whereas insufficient stock causes lost sales, customer dissatisfaction and weakened brand trust.

On-shelf availability (OSA) is therefore a central operating outcome. It refers to the extent to which a product that should be available for sale is physically present and purchasable on the selling shelf or designated display location. OSA is affected by forecasting, assortment planning, supplier reliability, inventory records, shelf replenishment, store processes and customer demand.

This chapter develops an integrated approach to assortment, stock measurement, inventory classification and replenishment. It also considers perishable merchandise, inventory accuracy, technology, governance and managerial trade-offs.

Learning Objectives

By the end of this chapter, the learner should be able to:

1. Explain assortment architecture, SKU management and the relationship between merchandise and inventory decisions.
2. Calculate and interpret stock turn, inventory levels, safety stock and reorder points.
3. Apply ABC analysis, economic order quantity and replenishment logic.
4. Diagnose the causes of poor on-shelf availability and inventory inaccuracy.
5. Evaluate the trade-offs among availability, working capital, markdowns, waste and shrinkage.
6. Design an operational response to a retail inventory problem using quantitative and managerial evidence.

1. Merchandise Architecture and SKU Decisions

1.1 Merchandise assortment

An assortment is the structured range of products offered by a retailer. It has several dimensions:

• Breadth: the number of product categories or product lines.
• Depth: the number of variants within a category, such as brands, sizes, colours or flavours.
• Price architecture: the distribution of entry-level, mainstream and premium products.
• Brand architecture: the balance among national brands, private labels and exclusive products.
• Seasonality: the timing and duration of product availability.
• Localization: adaptation of the assortment to regional demand, climate, culture and store mission.

A supermarket may have broad category coverage and substantial depth in staple categories. A convenience store typically has narrower breadth but emphasizes fast-moving products and immediate consumption. A fashion retailer may carry high depth in style and colour but face substantial demand uncertainty and obsolescence risk.

Assortment planning must distinguish customer value from operational complexity. Adding a SKU can improve choice and differentiation, but it also increases shelf-space requirements, ordering complexity, forecast error and the probability of slow-moving inventory. The relevant question is not whether a product sells, but whether its sales, margin, strategic role or customer value justify the resources it consumes.

1.2 SKU and item identity

A stock-keeping unit (SKU) is a uniquely managed item-location combination or item variant used for inventory control. A 500-gram cereal pack and a 1-kilogram pack are normally different SKUs. In omnichannel retail, the same product may require separate inventory records for a store, fulfilment centre and online channel.

A robust SKU master should include product description, barcode, unit of measure, dimensions, supplier, case pack, cost, retail price, tax status, shelf life, replenishment parameters and status. Poor master data can produce false availability: the system may show stock under an incorrect unit of measure or a discontinued product may remain active in the ordering system.

1.3 Product roles

Products can be classified by their role in the assortment:

A product’s role affects the appropriate service level. A retailer may tolerate occasional stockouts on a niche accessory but not on a high-frequency staple or a product advertised in a promotion.

2. Inventory Economics and Performance Measures

Inventory performs several functions: it decouples supply from demand, supports product variety, protects against uncertainty and enables economies of scale in purchasing and transportation. However, inventory also ties up working capital and creates operating risk.

2.1 Inventory components

Retail inventory may be divided into:

• Cycle stock: inventory expected to be consumed between replenishment deliveries.
• Safety stock: additional stock held against demand or supply uncertainty.
• Pipeline stock: inventory ordered but still in transit or processing.
• Presentation stock: units required to keep a shelf or display visually complete.
• Seasonal or speculative stock: inventory accumulated in anticipation of a predictable or strategic event.
• Dead stock: inventory with little realistic prospect of sale at the planned price.

The balance among these components should be explicit. A high inventory position is not necessarily healthy if it consists mainly of unwanted or inaccessible stock.

2.2 Inventory valuation and working capital

Inventory value is generally measured at cost for internal operational analysis, although accounting standards determine the formal financial treatment. A simplified inventory investment is:

Here, i denotes each SKU, units are physical or system-recorded units, and unit cost is the relevant acquisition or accounting cost. The measure should use consistent units and cost definitions. Retail selling price should not be substituted for cost when calculating investment.

2.3 Stock turn

Stock turn measures how many times average inventory is sold or consumed during a period. A common cost-based formula is:

Average inventory is often approximated as:

The formula assumes that opening and closing inventory reasonably represent the period’s average. Where inventory is highly seasonal or volatile, monthly or weekly averages are more reliable.

For example, if annual cost of goods sold is $1,200,000 and average inventory at cost is $300,000:

Approximate days of supply based on this turn rate are:

Using 365 days, the example gives 91.25 days. Turn should not be maximized blindly. Very high turn may indicate efficient capital use, but it may also indicate insufficient safety stock and frequent stockouts. Very low turn may be justified for strategic, seasonal or long-lead-time items, but otherwise signals excessive capital or weak demand.

2.4 Sell-through and gross margin return

Sell-through measures the proportion of available units sold during a defined period:

The formula must specify whether transfers, returns and markdown sales are included. For fashion and seasonal goods, sell-through at a particular date is often more informative than annual stock turn.

Gross margin return on inventory investment (GMROII) relates gross margin to the inventory investment required to earn it:

If gross margin is $240,000 and average inventory cost is $300,000, GMROII is 0.80. This means that each dollar of average inventory investment generated $0.80 of gross margin during the measured period. Definitions differ across organizations, so comparisons require consistent treatment of markdowns, shrinkage and inventory valuation.

3. Demand, Lead Time and Replenishment Logic

Replenishment converts demand information into purchase orders, transfers, allocation instructions or shelf-restocking tasks. The basic decision is how much to order and when to order it.

3.1 Lead time

Lead time is the elapsed time between initiating a replenishment decision and having usable stock available for sale. It may include order approval, supplier processing, transport, receiving, quality checks and put-away. A nominal supplier lead time is not sufficient if receiving delays or store processing are material.

Demand during lead time is estimated as:

If average demand is 40 units per day and lead time is 5 days, expected lead-time demand is 200 units. This calculation assumes relatively stable demand and lead time. When either is uncertain, safety stock is required.

3.2 Safety stock

Safety stock protects against uncertainty. For a simple model in which demand varies but lead time is constant, safety stock can be estimated as:

where z is the service-factor corresponding to the desired cycle service level, σ_d is the standard deviation of demand per time unit, and L is lead time in the same time unit. The formula assumes independent demand observations, approximately stable variability and constant lead time.

For example, suppose daily demand has a standard deviation of 12 units, lead time is 4 days and the desired service level uses z = 1.65:

If both demand and lead time vary, a commonly used independent-variability approximation is:

and:

Here, L̄ is average lead time, σ_d is standard deviation of demand per time unit, d̄ is average demand per time unit, and σ_L is standard deviation of lead time. This approximation assumes demand and lead time are independent. If they are correlated, a covariance term is required; therefore, the formula should not be applied mechanically.

Safety stock is not a substitute for correcting poor forecasts, unreliable suppliers or inaccurate inventory records. Increasing it may conceal process problems while consuming working capital.

3.3 Reorder point

The reorder point (ROP) is the inventory position at which replenishment should be initiated:

For the previous example, average demand is 40 units per day, lead time is 4 days and safety stock is 40 units:

An order is triggered when inventory position reaches 200 units. Inventory position is not simply shelf stock. It is normally calculated as:

The exact definition must reflect the retailer’s system. If 80 units are physically on hand and 140 units are on order, with no backorders or allocations, inventory position is 220 units and an order would not yet be triggered under the example policy.

3.4 Periodic and continuous review

Under continuous review, stock is monitored continuously or after every transaction, and an order is triggered when the reorder point is reached. This approach supports high-value or fast-moving products but requires reliable data and system capability.

Under periodic review, inventory is checked at fixed intervals. The target stock level for a review period plus lead time can be represented as:

The target level should cover expected demand over the review period plus lead time, together with appropriate safety stock. Periodic review is operationally convenient where suppliers visit on scheduled routes, but it exposes the retailer to additional demand uncertainty between reviews.

4. Economic Order Quantity and Its Limits

Economic order quantity (EOQ) identifies the order quantity that minimizes the sum of ordering and holding costs under a simplified set of assumptions:

where Q is order quantity in units, D is annual demand in units, S is ordering cost per order, and H is annual holding cost per unit. If holding cost is expressed as a percentage i of unit cost C, then H = i × C.

The classic EOQ model assumes relatively constant demand, constant lead time, replenishment in full batches, no quantity discounts, no shortages, known costs and negligible stockout effects. These assumptions are often unrealistic in retail, but EOQ remains useful as a benchmark.

Suppose annual demand is 10,000 units, ordering cost is $50 per order and annual holding cost is $2 per unit:

The retailer should not automatically order 707 units. Case-pack constraints, minimum order quantities, supplier discounts, shelf capacity, perishability and promotional calendars may produce a different practical quantity. EOQ also addresses order size, not the reorder point or service level.

5. ABC Analysis and Inventory Prioritisation

ABC analysis classifies SKUs according to annual consumption value:

A common policy is to rank SKUs from highest to lowest consumption value and calculate cumulative percentages. The category boundaries are managerial conventions rather than universal laws. A typical structure might designate approximately 70–80% of value as A items, the next 15–25% as B items and the remainder as C items.

ABC classification should not be confused with product importance. A low-cost item may be a traffic builder, legally required, essential to a product bundle or highly damaging to lose. A useful retail policy combines value-based ABC analysis with criticality, demand variability, margin, lead time and shrinkage risk.

For example, an ABC–XYZ approach may classify products by value and demand predictability: A-X items are high-value and predictable; A-Z items are high-value and erratic; C-X items are low-value and stable. A-Z products may require managerial judgment rather than automatic replenishment.

6. On-Shelf Availability and Inventory Accuracy

6.1 From system availability to customer availability

A system may report ten units on hand while the shelf is empty. The units may be in the back room, misplaced, damaged, reserved for an online order, awaiting processing or stolen. OSA therefore requires alignment among system inventory, physical inventory and shelf presentation.

A simple unit-based availability rate can be stated as:

The denominator must be carefully defined. It may refer to store-SKU observations, time intervals, planogram positions or customer demand occasions. A retailer should not compare OSA rates produced by different definitions.

A related metric is inventory record accuracy (IRA):

A record is “correct” only within a defined tolerance, such as an exact match or a permitted unit variance. High IRA is valuable, but it does not guarantee OSA if replenishment tasks are not executed.

6.2 Causes of stockouts

Stockouts arise from both physical and informational causes:

1. Forecast error or unexpected demand.
2. Supplier shortage, late delivery or poor fill rate.
3. Incorrect reorder parameters or case-pack settings.
4. Inventory-record errors caused by receiving, scanning, returns or transfers.
5. Stock hidden in the back room or stored in the wrong location.
6. Shelf capacity or planogram constraints.
7. Damaged, expired or quarantined units.
8. Shrinkage and unrecorded loss.
9. Delayed shelf replenishment or insufficient labour.
10. Promotional demand exceeding allocation.

The appropriate response depends on the cause. Increasing the reorder point will not solve a receiving error; ordering more will not solve theft; and improving the forecast will not solve a missed shelf-replenishment task.

6.3 Inventory counting systems

• Periodic physical inventory counts all or most inventory at a defined date.
• Cycle counting counts selected SKUs continuously throughout the year.
• Perpetual inventory updates records after transactions, usually through a point-of-sale and inventory-management system.
• Exception counting focuses on negative balances, large variances, high-value items or unusual movements.

A sound cycle-counting policy gives more frequent attention to A items, high-shrinkage items and products with repeated discrepancies. Variance investigation should identify root causes rather than simply adjusting the record.

7. Perishables, Markdown and Waste Management

Perishable products introduce a second dimension: inventory loses value with time even when it remains physically present. Replenishment must consider remaining shelf life, expiry dates, temperature control, first-expiry-first-out rotation and demand at the end of the selling window.

Markdowns can recover value from slow-moving or ageing stock, but excessive or late markdowns reduce margin and may train customers to delay purchases. A markdown decision should consider remaining demand, remaining selling days, inventory quantity, replacement cost, disposal cost, legal requirements and the effect on price image.

Waste differs from shrinkage. Waste is normally inventory discarded because of expiry, damage, spoilage or operational loss that is identified. Shrinkage is the unexplained difference between recorded and actual inventory, often associated with theft, administrative error or process failure. The distinction matters because prevention strategies differ.

For fresh goods, a retailer may use dynamic ordering, automated expiry alerts, reduced-order windows and electronic shelf labels. However, algorithms require accurate expiry, sales and inventory data. A system that optimizes on defective inputs can accelerate waste rather than reduce it.

8. Integrated Replenishment Governance

Inventory policies should be designed at the SKU–location level, because the same item may have different demand, shelf capacity and lead time in different stores. Governance should specify:

• ownership of forecasts and replenishment parameters;
• approval rules for overrides and manual orders;
• supplier service-level expectations;
• treatment of promotions and new-product launches;
• inventory adjustment authority;
• cycle-count frequency and tolerance levels;
• escalation procedures for stockouts and excess stock;
• audit trails for changes to master data.

Supplier performance is often measured using case fill rate, on-time delivery, lead-time reliability and invoice accuracy. A supplier with a nominally low price but poor fill rate can increase emergency freight, lost sales and safety-stock requirements.

Retailers should distinguish service level from fill rate. A cycle service level is the probability of not stocking out during a replenishment cycle, whereas fill rate is the proportion of demand supplied immediately. These measures are related but not identical; a single stockout can have a very different effect under each measure.

9. Technology and Artificial Intelligence

Retail inventory systems increasingly combine point-of-sale data, radio-frequency identification, electronic shelf labels, computer vision, mobile task management and machine-learning forecasts. These tools can improve transaction visibility, locate inventory and prioritize replenishment work.

Artificial intelligence is especially useful for detecting demand patterns, estimating promotion uplift, identifying anomalous inventory movements and recommending order quantities. Nevertheless, managerial controls remain necessary. Models can fail when products are new, promotions change consumer behaviour, supply is constrained, data is biased or store execution differs from historical patterns.

Responsible deployment requires data-quality monitoring, human override rules, explainable exception reports, privacy protection where customer data is used and controls against discriminatory allocation. Automation should support accountable decisions rather than obscure them.

10. Illustrative Case: FreshBite Urban Grocers

Background

FreshBite operates 18 urban supermarkets. Its fresh dairy category has strong sales but experiences frequent shelf stockouts in the afternoon and high evening waste. The category system reports an average of 2.5 days of inventory, while store managers believe the shelf is empty despite stock in the back room.

Problem

The category team initially proposes raising safety stock. However, an investigation shows three separate problems: demand is concentrated after work hours, deliveries are received late in the morning, and store staff do not consistently rotate stock or complete shelf-replenishment tasks. Inventory records also include units that have expired but were not written off.

Analysis

A four-week review shows average daily demand of 120 units for a high-volume yoghurt SKU. Demand standard deviation is 25 units per day, and effective lead time is two days. At a 95% cycle service level, use z = 1.65. Estimated safety stock is:

Expected lead-time demand is 240 units, so:

The calculation is only a starting point. The store’s shelf capacity is 180 units, making it impossible to hold the calculated reorder point entirely on the shelf. The policy therefore requires back-room availability, scheduled replenishment and reliable receiving rather than simply expanding shelf stock.

Approach and solution

FreshBite adopts the following measures:

1. It separates sellable, expired and quarantined units in the inventory system.
2. It schedules deliveries before the peak demand period where supplier capacity allows.
3. It introduces replenishment scans at 13:00, 16:00 and 18:00.
4. It uses first-expiry-first-out rotation and expiry-date alerts.
5. It counts the SKU daily for two weeks, then moves to risk-based cycle counting.
6. It measures OSA by time block rather than only once during the morning.
7. It reviews supplier fill rate and receiving delays weekly.
8. It sets a manager approval rule for safety-stock changes.

Outcome

After six weeks, afternoon OSA improves, expired-unit write-offs become visible and waste declines. The retailer does not eliminate all stockouts because demand remains variable, but it avoids the costly mistake of treating a shelf-execution problem as a pure purchasing problem.

Learning points

The case demonstrates that OSA is a system outcome. Quantitative parameters are useful only when inventory records, lead times, shelf processes and product condition are accurately represented. It also shows why availability, waste and working capital must be managed jointly.

Challenges and Limitations

Inventory models simplify reality. Demand may be intermittent, correlated with promotions or affected by competitor actions. Lead times may change abruptly. Supplier minimums and case packs can dominate the calculated order quantity. Forecast accuracy may deteriorate for new products. Safety stock based on historical variation can be inadequate during structural change.

Financial measures also require caution. High stock turn may result from understocking; high GMROII may reflect temporary price increases; and low inventory value may conceal severe availability problems. Metrics should therefore be reviewed together, with segmentation by product role and store type.

Ethically, retailers must avoid deliberate understocking of essential products merely to create scarcity or manipulate customers. Product traceability, expiry control and recall execution are consumer-safety responsibilities. Inventory adjustments must be auditable, and employee monitoring technologies should be proportionate and transparent.

Best-Practice Decision Framework

A practical sequence for managerial decisions is:

1. Define the customer and product service requirement.
2. Confirm SKU master data, unit of measure and shelf capacity.
3. Measure demand, variability, lead time and supplier performance.
4. Establish inventory position and identify record inaccuracies.
5. Segment SKUs using value, criticality, variability and perishability.
6. Select continuous or periodic review.
7. Set reorder points, safety stock and order quantities using appropriate assumptions.
8. Test feasibility against case packs, minimum orders, storage and cash limits.
9. Align receiving, back-room and shelf-replenishment processes.
10. Monitor OSA, stock turn, waste, shrinkage, markdowns and customer outcomes together.
11. Investigate exceptions and revise parameters through controlled governance.

Chapter Summary

Merchandise assortment determines the range and role of products offered, while inventory management determines how resources are committed to making those products available. SKU decisions must balance customer choice against operational complexity. Stock turn, sell-through and GMROII provide complementary views of inventory productivity, but none should be interpreted without considering service levels and product characteristics.

Safety stock protects against uncertainty, and the reorder point combines expected lead-time demand with that protection. EOQ provides a benchmark for order quantity under restrictive assumptions; it does not determine service level or replace managerial judgment. ABC analysis focuses attention on high-value items, but criticality, variability, perishability and shrinkage risk should also be considered.

On-shelf availability depends on accurate records, reliable supply, appropriate parameters and disciplined store execution. Waste, markdowns and shrinkage require different interventions. Technology and AI can improve visibility and forecasting, but only when supported by sound data governance, human oversight and ethical controls. Effective retail operations therefore manage availability, working capital and inventory risk as one integrated system.

Key Terms

Assortment: The structured range of products offered by a retailer.

SKU: A uniquely managed product variant or item-location inventory unit.

Breadth: The number of product categories or lines in an assortment.

Depth: The number of variants offered within a category.

Cycle stock: Inventory expected to be consumed between replenishment deliveries.

Safety stock: Additional inventory held to protect against uncertainty.

Lead time: Time from initiating replenishment to usable stock availability.

Reorder point: Inventory position at which replenishment is triggered.

Inventory position: On-hand stock adjusted for on-order, allocated and backordered quantities.

Stock turn: The number of times average inventory is sold or consumed during a period.

Sell-through: The proportion of available units sold during a specified period.

EOQ: Economic order quantity that minimizes simplified ordering and holding costs.

ABC analysis: Classification of items by annual consumption value.

On-shelf availability: The extent to which expected products are physically purchasable.

Inventory record accuracy: The degree to which recorded inventory agrees with physical inventory.

Markdown: A reduction in selling price intended to stimulate sale or recover value.

Waste: Identified inventory loss through expiry, spoilage, damage or disposal.

Shrinkage: Unexplained inventory loss or discrepancy between records and physical stock.

GMROII: Gross margin generated relative to average inventory investment.

Review Questions

Short Answer Questions

1. Distinguish assortment breadth from assortment depth.
2. What is the difference between cycle stock and safety stock?
3. Define the reorder point and state its basic components.
4. Why does high stock turn not always indicate superior performance?
5. Distinguish waste from shrinkage.

Long Answer Questions

1. Explain how assortment architecture affects inventory investment, customer choice and operational complexity.
2. Discuss the assumptions and managerial limitations of the EOQ model in retailing.
3. Explain how ABC analysis can be improved by adding criticality, variability and perishability.
4. Analyze the relationship among inventory accuracy, shelf replenishment and on-shelf availability.
5. Evaluate the role of technology and AI in retail replenishment, including governance risks.

Analytical / Application Questions

1. A SKU has average daily demand of 60 units, a demand standard deviation of 10 units, a four-day lead time and a service factor of 1.65. Assuming constant lead time, calculate safety stock and reorder point.
2. Annual demand is 24,000 units, ordering cost is $30 per order and annual holding cost is $1.50 per unit. Calculate EOQ and explain two reasons why the practical order quantity may differ.
3. A store reports 25 units on hand, 100 units on order, 10 units allocated to online orders and 5 units backordered. Calculate inventory position and explain whether a reorder point of 120 units has been reached.

Multiple Choice Questions with Answers

Question 1

Which assortment dimension refers to the number of variants within a product category?

A. Breadth

B. Depth

C. Criticality

D. Turnover

Correct Answer: B. Depth

Question 2

A retailer has annual cost of goods sold of $900,000 and average inventory at cost of $150,000. What is its stock turn?

A. 3 turns

B. 4 turns

C. 6 turns

D. 7.5 turns

Correct Answer: C. 6 turns

Question 3

If average daily demand is 50 units and lead time is 3 days, what is expected lead-time demand?

A. 53 units

B. 100 units

C. 150 units

D. 500 units

Correct Answer: C. 150 units

Question 4

Which formula represents the basic reorder point?

A. Annual demand ÷ unit cost

B. Lead-time demand + safety stock

C. Ordering cost + holding cost

D. Opening inventory − closing inventory

Correct Answer: B. Lead-time demand + safety stock

Question 5

In the simplified safety-stock formula, what does z represent?

A. Unit cost

B. Order quantity

C. Service-level factor

D. Supplier fill rate

Correct Answer: C. Service-level factor

Question 6

Annual demand is 10,000 units, ordering cost is $50 and annual holding cost is $2 per unit. The approximate EOQ is:

A. 50 units

B. 224 units

C. 500 units

D. 707 units

Correct Answer: D. 707 units

Question 7

ABC analysis normally classifies SKUs primarily according to:

A. Annual consumption value

B. Shelf colour

C. Supplier location only

D. Number of employees handling the SKU

Correct Answer: A. Annual consumption value

Question 8

A system reports stock on hand, but the shelf is empty because units are waiting in the back room. This is primarily an example of:

A. Excessive assortment breadth

B. A shelf-execution or location problem

C. A reduction in annual demand

D. An EOQ calculation error

Correct Answer: B. A shelf-execution or location problem

Question 9

Which measure compares gross margin with average inventory investment?

A. OSA

B. GMROII

C. Lead-time demand

D. Case pack

Correct Answer: B. GMROII

Question 10

Which statement best describes shrinkage?

A. Planned seasonal inventory

B. Identified expiry disposal only

C. Unexplained difference between recorded and actual inventory

D. Inventory ordered but still in transit

Correct Answer: C. Unexplained difference between recorded and actual inventory