In Part 1 of this series, we established that inventory exists because of uncertainty.
In Part 2, we explored how safety stock should be calculated using demand variability, lead time variability and the service level the business wants to achieve.
Part 3 addresses a question that follows naturally from both: should every product in the range receive the same inventory policy?
The answer is almost always no.
Not every product creates the same value, carries the same risk or deserves the same level of management attention. High-performing businesses recognise those differences, and manage their inventory accordingly.
Why Treating Every SKU the Same Is Expensive
Most businesses apply the same inventory logic to every product they sell.
The same safety stock assumptions. The same reorder rules. The same service level targets. The same cycle count frequency. Applied uniformly across hundreds or thousands of SKUs.
It is understandable. Uniform policies are simple to administer. And they work reasonably well when the product range is small and demand is predictable.
But as the range grows, uniform inventory policy becomes increasingly expensive.
Fast-moving high-value products may be under-stocked, creating stockouts, lost sales and customer service failures on the products that matter most.
Slow-moving low-value products may be over-stocked, absorbing working capital, storage space and management attention on products that contribute very little to revenue.
Both problems can exist simultaneously in the same business. Often without anyone noticing, because the overall inventory value looks acceptable even when the composition is wrong.
The solution isn't to manage inventory harder. It's to manage it differently across the product range.
ABC Classification: Where to Focus Management Effort
ABC classification applies the Pareto principle to inventory, the observation that a small number of products typically account for the majority of commercial results.
In most businesses the distribution follows a consistent pattern:
- A items (typically 10–20% of SKUs) account for around 70–80% of revenue or inventory value
- B items (typically around 20–30% of SKUs) account for around 15–20% of revenue or inventory value
- C items (typically 50–60% of SKUs) account for only around 5–10% of revenue or inventory value
The exact split varies by business and industry. The principle is consistent: a small number of products drive most of the commercial result.
ABC classification doesn't prescribe how to manage inventory. It identifies where to focus management effort.
A items, the products driving the most commercial value, generally justify closer attention, more frequent review and tighter stock control. C items can typically be managed with less intensity without materially affecting commercial performance. The right approach for each segment depends on factors beyond volume alone, including margin contribution, customer criticality, substitutability and supply risk.
Annual consumption value is a common starting point, while revenue, margin contribution or other commercial measures may also be relevant depending on the business. Some businesses add further segments when a very small number of products dominate results disproportionately. The goal is always the same: direct management effort where it creates the most value.
Adding Demand Variability: The ABC-XYZ Approach
ABC classification tells you which products are commercially important.
It doesn't tell you how predictable demand is, and that distinction matters significantly for inventory policy.
XYZ classification adds a second dimension by segmenting products according to demand variability:
- X items have stable, predictable demand, relatively straightforward to plan and manage
- Y items have variable demand, seasonal patterns, promotional effects or moderate fluctuation
- Z items have highly erratic or intermittent demand, difficult to forecast and plan reliably
When combined with ABC, this creates a matrix that gives a much more complete picture of what each product actually requires.
Consider the difference between:
Commercially important, predictable demand. Tight inventory control, calculated safety stock, frequent review.
Commercially important, but highly variable or intermittent demand. May require a fundamentally different planning approach, not simply more safety stock.
Lower commercial importance, predictable demand. Can be managed efficiently with simplified replenishment rules.
Lower commercial importance, erratic demand. Often the most expensive segment to manage poorly, tying up working capital in stock that moves unpredictably.
The key insight is that ABC classification alone can lead to the assumption that A items always need more stock and C items always need less. When demand variability is added, the picture becomes more nuanced, and more accurate.
For most NZ importers and distributors, ABC classification alone is a strong and practical starting point. XYZ analysis becomes increasingly valuable as the product range grows and demand patterns diverge across the range.
Service Levels and Commercial Priorities
In Part 2, we established that service level is a commercial decision, not a warehouse metric.
ABC classification provides a useful starting point for that conversation. Products that drive the most revenue and carry the greatest risk if unavailable generally justify a higher investment in availability. Products with lower commercial importance, easier substitutability and lower stockout cost may justify a more modest service level target.
One point worth reinforcing from Part 2: the relationship between service level and safety stock is not linear. Moving from 95% to 99% requires a disproportionately large increase in safety stock compared to moving from 80% to 95%. The cost of each additional percentage point of protection rises sharply as targets approach 100%.
This is why setting service level targets deliberately, by product segment, based on commercial priorities, is more effective than applying a single uniform target across the range. The right level for any product depends on its commercial importance, the cost of a stockout, the availability of substitutes and the business's overall inventory investment goals.
Reorder Points: When Should Replenishment Happen?
Once inventory has been segmented and service level decisions have been made, the next practical question is: when should the business place a replenishment order?
The reorder point answers that question. It defines the inventory level at which a new order should be triggered, early enough to cover demand during the replenishment lead time, with a safety stock buffer to protect against variability.
- Average daily demand: 20 units
- Lead time: 10 days
- Safety stock: 50 units
When stock reaches 250 units, a replenishment order should be placed.
The reorder point determines when to order. How much to order is a separate decision and should reflect the appropriate replenishment policy for that SKU, whether a fixed quantity, a variable amount based on current stock levels, or a more sophisticated approach that balances ordering frequency against holding costs.
Reorder points should be reviewed regularly. If average demand changes, if a supplier's lead time shortens or extends, or if safety stock is recalculated, the reorder point should be updated accordingly. Parameters set once and left unchanged become unreliable over time.
Min/Max Systems
A simpler alternative to calculated reorder points is the Min/Max system.
Min/Max defines two inventory thresholds. Min is the level at which a replenishment order is triggered. Max is the target inventory level after replenishment. The order quantity is the difference between current stock and the Max level, so both order timing and order quantity can vary depending on actual stock levels at the time of review.
Min/Max is widely used because it is simple to administer and easy to understand. It works well in environments where demand is relatively stable and predictable.
Its limitations become apparent when demand is highly variable, lead times fluctuate significantly or the product range is large and diverse. In those situations, replenishment parameters based on current demand, lead time and variability data generally produce more reliable results than static Min/Max settings.
A pragmatic approach works well for many NZ businesses: more precise replenishment methods for products where accuracy matters most, and simplified systems for products where ease of administration is more valuable than precision. The important thing is that parameters are reviewed regularly regardless of which method is used.
A Note on Intermittent Demand
For products with highly erratic or intermittent demand, the Z items in an ABC-XYZ framework, standard safety stock calculations and reorder point systems can become unreliable.
When a product sells zero units for several weeks and then receives a large unexpected order, average demand figures and standard deviation calculations may not reflect the actual demand pattern well enough to produce useful inventory parameters.
These products often require a different approach: reviewing them individually, considering whether to stock them at all, using safety lead time rather than safety stock in some cases, or applying specialist planning methods designed for intermittent demand patterns.
The detail of those approaches is beyond the scope of this series. What matters here is the recognition that Z items, particularly CZ items, deserve to be identified and treated differently, rather than having standard inventory policies applied to them by default.
Cycle Counting: Maintaining Inventory Accuracy
Inventory accuracy is the foundation of everything discussed in this series.
Safety stock calculations are only as reliable as the demand data that drives them. Reorder points only work if the system knows the correct on-hand quantity. Service level targets are meaningless if the inventory position is inaccurate.
The traditional approach to maintaining accuracy is an annual stocktake, counting everything once a year and reconciling the result. The problem is that errors accumulate throughout the year, decisions are made on inaccurate data for months, and the process itself is disruptive and time-consuming.
Cycle counting replaces the annual stocktake with a continuous counting programme. A small number of SKUs are counted regularly, rotating through the full product range over a defined period. Errors are identified and corrected in near-real time, making it far easier to identify root causes before they compound.
Products with greater value, transaction activity or operational risk generally justify more frequent counting. Products with lower commercial importance can be counted less frequently without materially increasing risk.
The key purpose of cycle counting isn't just to correct errors. It's to identify and eliminate their causes. A business that counts frequently and finds the same discrepancies repeatedly is treating symptoms rather than root causes. The discrepancies are signals about receiving processes, picking procedures or system transaction failures that deserve investigation.
The HSCM Inventory Policy Framework
Bringing these elements together, the following framework provides a practical guide to managing inventory differently across the product range.
Attention and control decrease from A items to C items. The framework guides where to focus, not a fixed rule for every SKU.
| Policy Area | A Items | B Items | C Items |
|---|---|---|---|
| Commercial importance | High | Medium | Lower |
| Service objective | Typically higher | Moderate | Typically lower |
| Management attention | High | Regular | Simplified |
| Replenishment method | More precise | Appropriate to SKU | Can be simplified |
| Parameter review | Frequent | Regular | Less frequent |
| Inventory control | Tighter | Standard | Simplified |
| Cycle counting | Frequent | Periodic | Less frequent |
ABC is only the starting point. Demand variability, criticality, supply risk and substitutability can all justify a different policy for individual SKUs, particularly when combined with XYZ analysis. The framework should guide decisions, not replace judgement.
Where to Start
If inventory is currently managed with a single uniform policy across the full range, the starting point isn't a major system change.
It's a set of practical questions:
- Have we classified our product range by commercial value, and when was that classification last reviewed?
- Are A items getting the management attention they deserve, and are C items consuming resources they don't justify?
- Do we understand which products have stable demand and which have variable or intermittent demand?
- Are we applying service level targets deliberately by product segment, or using a single uniform target across the range?
- Are our reorder points calculated from current demand and lead time data, or set manually and rarely updated?
- Are we cycle counting by commercial importance, or relying on an annual stocktake?
- Which products have demand patterns that standard inventory policies may not handle well?
These questions don't require new software. They require visibility of the data that already exists, and the discipline to use it differently across the product range.
Closing the Series
This series has followed a deliberate progression.
Part 1 established why inventory exists: not because businesses want to hold stock, but because uncertainty makes it necessary. Reduce the uncertainty, through better forecasting, more reliable suppliers, shorter lead times, and the inventory requirement reduces naturally.
Part 2 explored how to size the buffer correctly: using demand variability, lead time variability and service level targets to calculate safety stock rather than relying on rules of thumb and inherited assumptions.
Part 3 has addressed the final question: not every product deserves the same inventory policy. ABC classification, demand variability segmentation, differentiated service level decisions, appropriate replenishment methods and structured cycle counting all exist to direct management effort where it creates the most value.
Together, these three principles form a practical framework for inventory management that is both commercially grounded and operationally realistic.
Inventory management isn't about applying the same rules to everything. It's about focusing on what matters, protecting what's valuable, and simplifying what's low risk.
Not every SKU deserves the same attention. But every SKU deserves the right policy.