25 Costly Inventory Management Mistakes and the Research-Backed Fixes

Most inventory problems do not look like problems until the money is gone. The system says there are twelve units on the shelf, so nobody reorders. A planner adds "a bit extra, just in case" to every forecast. A warehouse keeps slow-moving stock for years because writing it off would hurt this quarter's numbers. Each decision feels sensible. Together they quietly drain cash, lose customers and leave the business exposed when the next disruption hits.

Inventory is one of the largest assets on many balance sheets, and one of the most misunderstood. Decades of research in operations management, from large field studies in retail stores to laboratory experiments with experienced managers, show the same costly mistakes again and again. The good news is that most of them have well-tested fixes.

We reviewed more than 60 peer-reviewed studies, official data series and reports from established publishers to identify the 25 most costly inventory management mistakes. Each one comes with the evidence, an APA citation you can check, and three practical fixes you can start using this week, whether you run a warehouse, a store network, a factory or a spare-parts operation.

The short version: Your inventory records are probably wrong, your stockouts cost more than you think, and your own judgement is biased in predictable ways. Fix the data first, measure what customers actually experience, use simple models with their limits in mind, and treat inventory as a deliberate buffer against risk, not just a cost to be cut.

Key takeaways

• Two out of three inventory records may be wrong. In a study of nearly 370,000 records across 37 stores of one retailer, about 65% did not match what was physically there (DeHoratius & Raman, 2008).
• Empty shelves are a chronic problem. Across 40 studies that reliably measured it, the average retail out-of-stock rate was 8.3% (Corsten & Gruen, 2003).
• A stockout costs more than the lost item. In a large catalogue field test, customers kept 86% of in-stock items they ordered but only 62% of out-of-stock items, and customers hit by stockouts on everything they ordered bought about 22% less afterwards (Anderson et al., 2006).
• Supply chain glitches hit profits hard. In the year before a glitch was announced, affected firms' operating income fell 107% relative to control firms (Hendricks & Singhal, 2005).
• Excess inventory punishes share prices. Excess inventory announcements were followed by an average two-day stock market reaction of about −6.8% (Hendricks & Singhal, 2009).
• Too lean is a real risk. In two-thirds of 54 manufacturing industries the inventory–performance link was significant, and it was mostly an inverted U: past a point, cutting inventory further hurt performance (Eroglu & Hofer, 2011).

How to read this guide: The 25 mistakes are grouped into four parts: bad data, stockouts, models and bias, and money and resilience. Use the table of contents to jump to the part you need. Citations in brackets lead to the full reference list at the bottom. This article is general information, not accounting, legal or engineering advice. For decisions with financial reporting or contractual consequences, involve your finance team and qualified advisers.

Part 1: Bad data: when the system lies

Mistake 1: Trusting the system count

Most inventory software assumes the number on the screen is the number on the shelf. Research says that assumption is usually false. DeHoratius and Raman (2008) examined nearly 370,000 inventory records from 37 stores of a single retailer and found about 65% were inaccurate. Product category explained around 26% of the variation in inaccuracy, far more than the individual store, which suggests the problem is built into processes and product types, not just a few sloppy sites. Earlier field work by Raman et al. (2001) found the same 65% figure at one leading retailer and estimated that execution problems like these reduce profits by more than 10%.

How to fix it:

• Measure record accuracy directly: count a random sample of locations each week and track the percentage that match the system within an agreed tolerance.
• Report accuracy by product category, not just by site. The research suggests category drives most of the variation.
• Treat accuracy as a KPI on the same dashboard as service level and inventory value, so it gets management attention.

Mistake 2: Missing the phantom stockout

The most dangerous record error is when the system shows stock but the shelf is empty. Automatic replenishment sees a positive balance and never reorders, so the item stays out of stock indefinitely. DeHoratius et al. (2008) call this "freezing" and showed that a replenishment tool which treats the true stock level as uncertain, updated with sales data, recovers much of the cost of record inaccuracy. Kang and Gershwin (2005) modelled how even a small rate of unrecorded stock loss, such as theft or damage, can corrupt records enough to cause serious stockouts. The revenue lost can exceed the value of the stock that disappeared, and lean systems with little slack are the most sensitive.

How to fix it:

• Flag every item that has positive system stock but zero sales for longer than its normal selling interval, and check it physically.
• Build "no sales despite stock" alerts into your replenishment system, ranked by expected lost sales.
• When an item is confirmed missing, correct the record the same day and investigate the cause, not just the number.

Mistake 3: Relying on one big annual count

Many organisations still treat a once-a-year wall-to-wall stocktake as their main accuracy control. By the time errors are found, they have been distorting orders for months. DeHoratius and Raman (2008) found that auditing practices were among the factors associated with lower record inaccuracy. Kök and Shang (2007) showed that the best policy links counting to replenishment: inspect when the recorded stock falls below a threshold, and adjust the order-up-to level depending on how long it has been since the last count. They used this structure to design a practical cycle-count programme that performed almost as well as the optimal policy.

How to fix it:

• Move to continuous cycle counting, with high-value and fast-moving items counted most often.
• Trigger extra counts on events, such as a recorded balance hitting zero or a low threshold, a picking shortage, or a customer complaint.
• Use the results to target causes: if one zone or category keeps failing, fix the process there rather than counting harder everywhere.

Mistake 4: Treating shrink as a cost of doing business

Shrink is the gap between what records say you should have and what you actually have. An industry survey by the National Retail Federation (2023) put the average U.S. retail shrink rate at 1.6% in fiscal year 2022, equal to $112.1 billion in losses, with internal and external theft together making up about 65% of it. Shrink matters beyond the direct loss, because it is a major source of record errors. Chuang and Oliva (2015) combined simulation with store data and found that backroom and shelf shrinkage were the predominant drivers of inventory record inaccuracy, while other recording and shelving errors had negligible effects in their model.

How to fix it:

• Record every known loss, damage or write-off at the time it happens so the system stays aligned with reality.
• Analyse shrink by item, location and shift to find patterns. Shrink is rarely spread evenly.
• Secure the backroom and receiving areas as carefully as the sales floor. That is where much unrecorded loss happens.

Mistake 5: Blaming people instead of fixing the process

When counts are wrong, it is tempting to blame "careless staff". The evidence points to staffing and process design. Chuang and Oliva (2015) found that full-time labour reduced inventory record inaccuracy, while part-time labour did not. They also identified a reinforcing loop: poor records create extra work, which squeezes the labour available to keep records right. Ton and Huckman (2008) studied 48 months of data from stores of a major U.S. retail chain. Employee turnover was associated with lower profit margins and customer service, but not in stores where managers enforced standard procedures consistently. Process discipline protected performance even when people left.

How to fix it:

• Write simple standard procedures for receiving, put-away, picking and adjustments, and train every new starter on them in week one.
• Check that critical inventory tasks are covered by experienced, stable staff rather than left to whoever is free.
• When an error occurs, ask "which step allowed this?" before asking "who did this?"

Mistake 6: Expecting technology to fix everything at once

Technology can help, but its value depends on where you use it. Hardgrave et al. (2013) ran two field experiments with a major global retailer. In the first, daily counts in 13 stores over 23 weeks showed that RFID-enabled visibility cut inventory record inaccuracy by about 26%. In the second, across 62 stores and five categories, the benefit ranged from no significant improvement to 81%, depending on the category. RFID worked best where the known drivers of inaccuracy were strongest. Kök and Shang (2007) make a related point: accurate information has a measurable value, and that value should be compared with its cost.

How to fix it:

• Pilot new tracking technology in the categories with the worst accuracy and highest value first, with a control group for comparison.
• Fix basic processes such as receiving checks and location discipline before automating them. Technology amplifies whatever process it sits on.
• Build a business case on measured accuracy gains and avoided stockouts, not vendor estimates.

Mistake 7: Measuring availability in the warehouse, not where customers shop

Many companies report high fill rates from their distribution centres while customers face empty shelves. Corsten and Gruen (2003), drawing on a worldwide study of out-of-stocks in fast-moving consumer goods (Gruen et al., 2002), found an average out-of-stock rate of 8.3% across 40 studies that reliably measured it, with little evidence of improvement over decades. Their conclusion was that the problem must be tackled on the shop floor, where ordering and replenishment practices break down. Ton and Raman (2010) found a related trap: "phantom products" that are in the store but stuck in storage areas where customers cannot find them. More variety and more inventory increased these phantom products, which reduced sales.

How to fix it:

• Measure on-shelf or point-of-use availability, not just warehouse fill rate. Use audits, shelf scans or sales-gap analytics.
• Track how long stock sits in back rooms or staging areas before it reaches the point of sale or use.
• Make store or site replenishment teams accountable for availability, and give them the time and tools to deliver it.

Part 2: Stockouts and the real cost of empty shelves

Mistake 8: Counting a stockout as one lost sale

Most inventory models treat a stockout as a single lost unit. Anderson et al. (2006) measured what really happens in a large field test with a mail-order catalogue company. Customers kept 86% of the in-stock items they ordered, but only 62% of items that were out of stock when ordered. The damage spread to the rest of the order: when one item was unavailable, the conversion rate on the other items in the same order fell from 83.1% to 71.2%. The effect also lasted. Customers for whom every item ordered was out of stock had about 22% lower demand afterwards than customers who had no stockouts.

How to fix it:

• Estimate stockout cost as lost margin on the item, plus lost margin on the basket, plus a share of future customer value.
• Use that fuller cost when setting service levels for key items. It usually justifies higher availability on the items customers care most about.
• Track repeat-purchase rates for customers who experienced stockouts, so the long-run cost is visible.

Mistake 9: Assuming customers will simply wait or switch brands

Many managers believe a stockout just shifts demand to another brand or another day. Corsten and Gruen (2003) found that shoppers increasingly switch stores when they cannot find what they want, and may never come back. Laboratory research shows why the reaction can be so strong. Fitzsimons (2000) ran four experiments and found that the more committed a shopper was to the missing option, the lower their satisfaction with the decision and the more likely they were to switch stores on later trips. Corsten and Gruen (2004) summarised the problem for managers in a single phrase: stock-outs cause walkouts.

How to fix it:

• Identify your "destination" items, the ones customers come specifically to buy, and protect their availability first.
• Offer a clear, close substitute or reliable delivery date when an item is out, rather than leaving a gap.
• Survey or sample customers who met a stockout to learn what they actually did next.

Mistake 10: Handling stockouts with discounts instead of honesty

When an item is unavailable, many firms reflexively offer a discount to keep the order. Anderson et al. (2006) tested five scripts with customers who ordered an out-of-stock item. Explaining that the item was out of stock because it was extremely popular kept the most orders, with 68% shipped and kept compared with 60% under the standard response. A 10% discount also retained orders in the short run, but in the long run discounts to encourage back-orders were the least profitable response the authors evaluated. The authors caution that the "popular" explanation only works, and is only ethical, when it is true.

How to fix it:

• Train customer-facing staff to give a truthful reason and a realistic delivery date for every out-of-stock item.
• Test your stockout responses with real customers before rolling out blanket discounts.
• Track long-run customer behaviour, not just whether the immediate order was saved.

Mistake 11: Adding variety without counting its operational cost

New products, sizes and variants can lift sales. They also spread inventory thinner and make execution harder. Wan et al. (2012) studied three years of weekly data from 108 distribution centres of a major soft-drink bottler. More product variety was linked to lower fill rates, and beyond a certain point additional variety actually reduced total sales: "too much of a good thing". Ton and Raman (2010) found that higher variety and inventory levels increased phantom products and defects in stores, offsetting some of the sales benefit.

How to fix it:

• Before adding a new SKU, estimate its effect on fill rate, handling and space, not just its sales potential.
• Review the long tail regularly. Remove or consolidate items that add complexity without adding meaningful sales or margin.
• Agree a "one in, one out" rule for categories where space or picking capacity is tight.

Mistake 12: Underestimating what a supply glitch does to profit

Production or shipment problems are often treated as operational hiccups. Hendricks and Singhal (2003) analysed 519 announced supply chain glitches and found an average abnormal fall in shareholder value of 10.28%. In a follow-up study of 885 glitches, affected firms saw operating income fall 107% relative to similar control firms in the year leading up to the announcement. They also had 6.92% lower sales growth, 10.66% higher cost growth and 13.88% higher inventory growth, with no recovery over the following two years (Hendricks & Singhal, 2005). It did not matter who caused the glitch or what industry the firm was in.

How to fix it:

• Map your critical items and single points of failure, and agree in advance what you will do if each one fails.
• Monitor early warning signals such as supplier delivery performance, lead-time drift and quality holds.
• Present glitch risk to senior leaders in financial terms, using evidence like this, so prevention gets funded.

Part 3: Models, maths and human bias

Mistake 13: Running ABC on sales value alone

ABC classification, sorting items into A, B and C classes by annual sales value, is one of the most widely used inventory tools. It is useful for focusing attention, but it is often misused to set service levels. Teunter et al. (2010) tested it on three large real-life datasets. Ranking by demand value or demand volume, with one fixed service level per class, produced solutions far from cost optimal. A criterion that also considers each item's holding cost, order quantity and shortage cost performed much better, and could take item criticality into account. Flores and Whybark (1986) had already shown that a single criterion struggles when managers must weigh more than one important dimension, such as criticality or lead time.

How to fix it:

• Use ABC to decide where to spend management attention, but set service levels using cost and criticality, not class alone.
• Add at least one more dimension, such as criticality to customers or production, or supply lead-time risk.
• Re-run the classification at least quarterly. Items move between classes as demand changes.

Mistake 14: Forecasting lumpy demand like smooth demand

Many items, especially spare parts, sell sporadically: nothing for weeks, then a large order. Standard forecasting methods handle this badly. Croston (1972) showed that exponential smoothing produces biased forecasts and inappropriate stock levels for intermittent demand, and proposed forecasting demand size and the interval between demands separately. Syntetos and Boylan (2005) tested methods on 3,000 real intermittent demand series from the automotive industry, and their improved Croston-type method performed best. Syntetos et al. (2005) proposed classifying demand by how often it occurs and how variable its size is, the basis of many "XYZ"-style schemes, so that each pattern gets a suitable method.

How to fix it:

• Classify every item by demand pattern (smooth, intermittent, erratic or lumpy) as well as by value.
• Use methods designed for intermittent demand on sporadic items instead of moving averages or simple smoothing.
• Combine ABC and pattern classes in a simple matrix, and set a policy for each cell rather than one rule for everything.

Mistake 15: Treating EOQ as gospel

The economic order quantity (EOQ) formula, which balances ordering costs against holding costs, was first published by Ford Whitman Harris in 1913 (Harris, 1913/1990; Erlenkotter, 1990). It assumes steady, known demand, fixed costs and instant replenishment, which are rarely true. Its strength is that total cost is flat near the optimum: by the formula's own maths, ordering twice the EOQ raises the combined ordering and holding cost by only about 25%. Its weakness is that it says nothing about uncertainty. Rumyantsev and Netessine (2007) studied 722 U.S. public companies from 1992 to 2002 and found that real inventories follow what classical models predict in several ways, rising with demand uncertainty, lead times and gross margins, but the evidence on holding costs was mixed.

How to fix it:

• Use EOQ as a starting point for order size, then adjust for minimum order quantities, pallet sizes, shelf life and capacity.
• Do not spend weeks perfecting order-cost estimates. The flat cost curve means a reasonable estimate is good enough.
• Handle uncertainty separately through safety stock and reorder points. EOQ does not do that job.

Mistake 16: Setting safety stock by gut feel

Safety stock is often set as "two weeks of cover" for everything, regardless of how uncertain demand or supply is. Standard textbooks show that safety stock should depend on demand variability, lead time and its variability, and the service level you want (Silver et al., 2017). The maths has consequences. With normally distributed demand, raising the cycle service level from 95% to 99% raises the required safety stock by about 41%, because the safety factor rises from 1.645 to 2.326. Location matters too. Eppen (1979) showed that pooling demand from several locations reduces costs, and that with identical, independent demands, inventory costs grow only with the square root of the number of locations combined. Graves and Willems (2000) developed a model for placing safety stock across multi-stage supply chains, which product teams at Eastman Kodak used to reduce finished-goods inventory.

How to fix it:

• Calculate safety stock per item from forecast error and lead-time data, not a blanket "weeks of cover" rule.
• Be clear which service measure you use, cycle service level or fill rate, and set targets by item importance.
• Look for pooling opportunities: holding slow movers centrally can cut total stock while keeping service.

Mistake 17: Measuring forecast accuracy with the wrong yardstick

Many teams judge forecasts with a single percentage error number. Hyndman and Koehler (2006) showed that many common accuracy measures are inadequate and break down in ordinary situations. Percentage errors, for example, become infinite or undefined when actual demand is zero, which is common for slow movers. They proposed the mean absolute scaled error as a standard measure for comparing accuracy across many series. Method choice matters too. The M4 forecasting competition compared 61 methods on 100,000 time series and found that combining methods generally beat relying on a single one (Makridakis et al., 2020).

How to fix it:

• Track both bias (are we consistently high or low?) and accuracy, using measures that work with zero demand.
• Measure error at the level and time horizon that drives decisions, such as item-location at the replenishment lead time.
• Test simple combinations of forecasting methods before investing in complex single models.

Mistake 18: Fiddling with the forecast

Planners often adjust system forecasts by small amounts based on instinct. Fildes et al. (2009) studied more than 60,000 forecasts and outcomes from four supply chain companies. Adjustments improved accuracy on average in three of the four, but the pattern mattered. Large adjustments tended to help, while small adjustments often made accuracy worse. Upward adjustments were much less likely to improve accuracy than downward ones and were more often in the wrong direction, which the authors read as optimism bias.

How to fix it:

• Only override a statistical forecast when you have specific information the model cannot see, such as a confirmed promotion or a lost customer.
• Require a short written reason for every override, then review which kinds of overrides actually improved accuracy.
• Be especially sceptical of small upward tweaks. The evidence says they rarely help.

Mistake 19: Ordering toward the average (the pull-to-centre bias)

For single-period decisions, such as seasonal goods, fresh food or event stock, the "newsvendor" model gives the profit-maximising order quantity. People consistently miss it. Schweitzer and Cachon (2000) found in two experiments that participants ordered too few high-profit products and too many low-profit products, pulling their orders toward average demand. Feedback and training did not remove the bias. Bolton and Katok (2008) found that how experience and feedback are organised can influence whether people learn to order closer to the optimum. Moritz et al. (2013) tested experienced supply chain professionals and students. People with higher cognitive reflection were less likely to chase last period's demand, and cognitive reflection predicted performance better than years of experience or managerial position.

How to fix it:

• Calculate the critical ratio for each item: the cost of having too little divided by the combined cost of too little and too much. Order to that point, not to the average.
• For high-margin items, expect to end the season with some leftovers. That is usually the profitable outcome, not a failure.
• Use decision rules or tools for routine orders, and do not let last period's demand drive the next order by itself.

Mistake 20: Overreacting to orders and feeding the bullwhip

The bullwhip effect is the tendency for order swings to grow as they move upstream from customers to suppliers. Lee et al. (1997a, 1997b) described it at Procter & Gamble, where modest swings in retail sales of Pampers became much bigger swings in orders further up the chain, and identified four causes: demand signal processing, rationing and shortage gaming, order batching and price promotions. Sterman (1989) showed in the classic "beer game" experiment that people ignore orders they have already placed but not yet received, and Croson and Donohue (2006) found the bullwhip persists even when the usual operational causes are removed. In real company data, about two-thirds of 4,689 U.S. public firms showed bullwhip between 1974 and 2008, although its average size fell by a third in the later period (Bray & Mendelson, 2012).

How to fix it:

• Always include stock already on order when deciding the next order, not just stock on hand.
• Share point-of-sale or consumption data with suppliers, and ask customers for theirs.
• Reduce batching and price-driven forward buying. Smaller, more frequent orders and stable pricing calm the whole chain.

Part 4: Money, strategy and resilience

Mistake 21: Underestimating what it costs to carry inventory

Many businesses count only the purchase price and maybe interest when they think about inventory cost. The full carrying cost also includes storage, handling, insurance, damage, shrink and the risk of obsolescence. Azzi et al. (2014) studied ten companies and found that managers computed holding costs in different ways, and that the type of storage system, manual forklift operations versus automated storage and retrieval, changed the cost structure. Firms that carry too much pay in other ways too. Chen et al. (2005) found that U.S. manufacturers with abnormally high inventories had abnormally poor long-term stock returns. Capkun et al. (2009) found a significant positive link between better inventory performance and financial performance across U.S. manufacturing from 1980 to 2005.

How to fix it:

• Build a full holding-cost rate with your finance team, covering capital, space, handling, insurance, shrink and obsolescence.
• Review the rate at least once a year. It changes with interest rates, warehouse costs and product life cycles.
• Use the same rate in every decision that trades inventory against other costs, from order sizes to supplier offers.

Mistake 22: Letting obsolete stock sit until it becomes a write-off

Slow-moving stock rarely gets cheaper to deal with. Under international accounting standards, inventory must be measured at the lower of cost and net realisable value, so a fall in what stock can be sold for has to be recognised as an expense (IFRS Foundation, n.d.). Song and Zipkin (1996) showed that the prospect of obsolescence should substantially change how inventory is managed, and that the option to dispose of excess stock makes the problem more manageable. Markets notice the mistakes. Hendricks and Singhal (2009) studied 276 excess inventory announcements and found a mean two-day stock market reaction of about −6.8% to −6.9%. Kesavan and Mani (2013) found an inverted-U relationship between retailers' abnormal inventory growth and their earnings a year later.

How to fix it:

• Run a monthly ageing report showing stock with no movement for 90, 180 and 365 days, by value.
• Agree clear triggers and owners for action: return to supplier, transfer, promote, rework, donate or dispose.
• Build end-of-life plans into new product launches, so run-out of the old item is managed from the start.

Mistake 23: Benchmarking inventory turns without context

Inventory turnover is a popular benchmark, but comparing raw turns across businesses can mislead. Gaur et al. (2005) studied 311 U.S. public retailers from 1987 to 2000. Turnover was linked to gross margin, capital intensity and sales surprise, the gap between actual and expected sales. Their model explained 66.7% of the variation within firms. They proposed an adjusted turnover measure that controls for these factors. Over the long run, Chen et al. (2005) found median inventory holding periods for U.S. manufacturers fell from 96 to 81 days between 1981 and 2000, mainly through work-in-process cuts, while finished goods did not decline. Inventory productivity still matters to investors: Alan et al. (2014) found that U.S. retailers with higher inventory turnover had better future stock returns.

How to fix it:

• Compare turns against your own history and close peers with similar margins and business models, not a generic industry figure.
• Break turns down by inventory type (raw materials, work in process, finished goods) and by product family.
• Pair turns with service level and obsolescence. High turns with empty shelves is not success.

Mistake 24: Confusing lean with "as little inventory as possible"

Lean and just-in-time thinking, rooted in the Toyota Production System, treats overproduction and excess stock as waste (Ohno, 1988). Done well, it pays. Shah and Ward (2003) found that bundles of lean practices explained about 23% of the variation in plants' operational performance. But lean does not mean zero buffers. Hopp and Spearman (2004) argue that the essence of pull is limiting work in process, not eliminating inventory. Eroglu and Hofer (2011) found the inventory–performance relationship was significant in two-thirds of 54 industries, and usually concave: beyond an optimal level of leanness, performance got worse.

COVID-19 exposed the cost of stripping buffers out. U.S. Census Bureau data show the total business inventories-to-sales ratio jumped to 1.74 in April 2020 as sales collapsed, then fell to a low of 1.26 in April 2021 as demand recovered faster than supply (U.S. Census Bureau, 2026). Ivanov (2020) showed that epidemic outbreaks are a distinct kind of supply chain risk that ripples across many stages at once. A McKinsey Global Institute (2020) analysis, a consulting estimate, suggested companies can expect disruptions lasting a month or longer every 3.7 years on average, with losses over a decade worth more than 40% of a year's profits. Shih (2020) argued that the pandemic forces firms to rethink lean, minimal-inventory global supply chains and make their networks more resilient.

How to fix it:

• Keep lean's discipline of exposing waste, but hold deliberate strategic buffers for critical, long-lead or single-source items.
• Stress-test your supply chain: ask what happens if a key supplier, port or plant stops for four to eight weeks (Chopra & Sodhi, 2004).
• Review buffer policies after every major disruption, and agree in advance how they will be funded.

Mistake 25: Running every product through the same supply chain

One inventory policy for every product is simple, and usually wrong. Fisher (1997) argued that functional products with predictable demand need efficient, low-cost supply chains, while innovative products with uncertain demand need responsive chains that keep buffers and capacity to react. Lee (2002) extended this to supply uncertainty, describing efficient, responsive, risk-hedging and agile strategies. The payoff from matching can be large. Fisher and Raman (1996) helped a fashion skiwear firm use early sales to guide later production, and costs fell enough to increase profits by 60% compared with its existing informal approach. Cachon and Fisher (2000) found that sharing demand and inventory data lowered supply chain costs by 2.2% on average, while halving lead times cut costs by about 21% and halving batch sizes by about 22%. Speeding up the physical flow was worth far more than extra information. Fisher and Raman (2010) later turned lessons like these into a practical playbook for using sales data to plan assortments and inventory.

How to fix it:

• Segment products by demand predictability and supply risk, and design a different inventory policy for each segment.
• For uncertain, short-life products, invest in shorter lead times and the ability to react to early sales.
• For stable products, focus on efficiency, lean flows and low cost, but keep buffers where supply risk is high.

Your 10-point inventory action plan

You do not need a new system to make progress. Across all 25 mistakes, the same practical moves appear again and again:

1. Measure record accuracy. Count a sample every week and report accuracy alongside service and inventory value.
2. Hunt phantom stockouts. Flag items with stock on the system but no sales, and check them physically.
3. Cycle count by risk. Count high-value, fast-moving and error-prone items most often.
4. Measure what customers see. Track availability on the shelf or at the point of use, not just in the warehouse.
5. Price stockouts properly. Include lost baskets and future customer value, not just the missing unit.
6. Classify on more than value. Combine ABC with demand pattern and criticality, and set policies by segment.
7. Calculate safety stock. Base it on forecast error, lead time and a chosen service level, not weeks of cover.
8. Discipline your overrides. Adjust forecasts only with specific new information, and review the results.
9. Manage the tail and the end of life. Review ageing stock monthly and act before it becomes a write-off.
10. Buffer on purpose. Keep lean flows, but hold deliberate stock where supply risk and customer impact are highest.

Want to build these skills in your team? Explore BLTI's logistics, transportation and operations credentials to strengthen inventory, planning and supply chain leadership, and take our free Leadership Flex Check to see how you lead when the pressure is on.

Frequently asked questions

What are the most common inventory management mistakes?

The most common and costly mistakes are trusting inaccurate system records, ignoring phantom stockouts, measuring availability in the warehouse rather than where customers shop, setting safety stock by gut feel, overriding forecasts without good reason, and letting slow-moving stock age into write-offs. Research suggests record inaccuracy alone affects a majority of records at some retailers (DeHoratius & Raman, 2008).

What is a phantom stockout?

A phantom stockout happens when the inventory system shows stock that is not physically available. Because the system thinks stock exists, it does not reorder, so the item can stay unavailable for a long time. Bayesian replenishment tools and targeted audits can reduce the problem (DeHoratius et al., 2008).

How much safety stock should I hold?

There is no single right answer. Safety stock should reflect demand variability, lead time and its variability, and the service level you want for each item (Silver et al., 2017). Higher service levels require disproportionately more stock, so set targets by item importance rather than using one blanket rule.

Is the EOQ formula still useful?

Yes, as a starting point. EOQ balances ordering and holding costs, and total cost is flat near the optimum, so rough cost estimates are usually good enough. It does not handle uncertainty, minimum order quantities or capacity limits, so pair it with safety stock calculations and practical adjustments.

Did COVID-19 prove that lean inventory is a mistake?

No. It showed that lean taken to extremes is a mistake. Lean practices are linked to better operational performance (Shah & Ward, 2003), but research finds an optimal level of leanness beyond which performance falls (Eroglu & Hofer, 2011). The lesson is to keep lean discipline while holding deliberate buffers for critical and high-risk items.

About this research

This guide was compiled by the Better Leaders Training Institute (BLTI) research team. Every source below was checked for authors, year, title, publisher and DOI or stable link, using publisher records and the Crossref database. We prioritised peer-reviewed studies in operations management, official government data and books by recognised experts. Where a figure comes from an industry survey or a consulting analysis, we say so in the text. Findings from individual studies describe specific samples and settings; your organisation may differ. This article is general information, not accounting, legal, financial or engineering advice.

References

Alan, Y., Gao, G. P., & Gaur, V. (2014). Does inventory productivity predict future stock returns? A retailing industry perspective. Management Science, 60(10), 2416–2434. https://doi.org/10.1287/mnsc.2014.1897

Anderson, E. T., Fitzsimons, G. J., & Simester, D. (2006). Measuring and mitigating the costs of stockouts. Management Science, 52(11), 1751–1763. https://doi.org/10.1287/mnsc.1060.0577

Azzi, A., Battini, D., Faccio, M., Persona, A., & Sgarbossa, F. (2014). Inventory holding costs measurement: A multi-case study. The International Journal of Logistics Management, 25(1), 109–132. https://doi.org/10.1108/IJLM-01-2012-0004

Bolton, G. E., & Katok, E. (2008). Learning by doing in the newsvendor problem: A laboratory investigation of the role of experience and feedback. Manufacturing & Service Operations Management, 10(3), 519–538. https://doi.org/10.1287/msom.1060.0190

Bray, R. L., & Mendelson, H. (2012). Information transmission and the bullwhip effect: An empirical investigation. Management Science, 58(5), 860–875. https://doi.org/10.1287/mnsc.1110.1467

Cachon, G. P., & Fisher, M. (2000). Supply chain inventory management and the value of shared information. Management Science, 46(8), 1032–1048. https://doi.org/10.1287/mnsc.46.8.1032.12029

Capkun, V., Hameri, A.-P., & Weiss, L. A. (2009). On the relationship between inventory and financial performance in manufacturing companies. International Journal of Operations & Production Management, 29(8), 789–806. https://doi.org/10.1108/01443570910977698

Chen, H., Frank, M. Z., & Wu, O. Q. (2005). What actually happened to the inventories of American companies between 1981 and 2000? Management Science, 51(7), 1015–1031. https://doi.org/10.1287/mnsc.1050.0368

Chopra, S., & Sodhi, M. S. (2004). Managing risk to avoid supply-chain breakdown. MIT Sloan Management Review, 46(1), 53–61.

Chuang, H. H.-C., & Oliva, R. (2015). Inventory record inaccuracy: Causes and labor effects. Journal of Operations Management, 39–40, 63–78. https://doi.org/10.1016/j.jom.2015.07.006

Corsten, D., & Gruen, T. (2003). Desperately seeking shelf availability: An examination of the extent, the causes, and the efforts to address retail out-of-stocks. International Journal of Retail & Distribution Management, 31(12), 605–617. https://doi.org/10.1108/09590550310507731

Corsten, D., & Gruen, T. (2004). Stock-outs cause walkouts. Harvard Business Review, 82(5). https://hbr.org/2004/05/stock-outs-cause-walkouts

Croson, R., & Donohue, K. (2006). Behavioral causes of the bullwhip effect and the observed value of inventory information. Management Science, 52(3), 323–336. https://doi.org/10.1287/mnsc.1050.0436

Croston, J. D. (1972). Forecasting and stock control for intermittent demands. Operational Research Quarterly, 23(3), 289–303. https://doi.org/10.1057/jors.1972.50

DeHoratius, N., Mersereau, A. J., & Schrage, L. (2008). Retail inventory management when records are inaccurate. Manufacturing & Service Operations Management, 10(2), 257–277. https://doi.org/10.1287/msom.1070.0203

DeHoratius, N., & Raman, A. (2008). Inventory record inaccuracy: An empirical analysis. Management Science, 54(4), 627–641. https://doi.org/10.1287/mnsc.1070.0789

Eppen, G. D. (1979). Effects of centralization on expected costs in a multi-location newsboy problem. Management Science, 25(5), 498–501. https://doi.org/10.1287/mnsc.25.5.498

Erlenkotter, D. (1990). Ford Whitman Harris and the economic order quantity model. Operations Research, 38(6), 937–946. https://doi.org/10.1287/opre.38.6.937

Eroglu, C., & Hofer, C. (2011). Lean, leaner, too lean? The inventory-performance link revisited. Journal of Operations Management, 29(4), 356–369. https://doi.org/10.1016/j.jom.2010.05.002

Fildes, R., Goodwin, P., Lawrence, M., & Nikolopoulos, K. (2009). Effective forecasting and judgmental adjustments: An empirical evaluation and strategies for improvement in supply-chain planning. International Journal of Forecasting, 25(1), 3–23. https://doi.org/10.1016/j.ijforecast.2008.11.010

Fisher, M. L. (1997). What is the right supply chain for your product? Harvard Business Review, 75(2), 105–116. https://hbr.org/1997/03/what-is-the-right-supply-chain-for-your-product

Fisher, M., & Raman, A. (1996). Reducing the cost of demand uncertainty through accurate response to early sales. Operations Research, 44(1), 87–99. https://doi.org/10.1287/opre.44.1.87

Fisher, M., & Raman, A. (2010). The new science of retailing: How analytics are transforming the supply chain and improving performance. Harvard Business Press.

Fitzsimons, G. J. (2000). Consumer response to stockouts. Journal of Consumer Research, 27(2), 249–266. https://doi.org/10.1086/314323

Flores, B. E., & Whybark, D. C. (1986). Multiple criteria ABC analysis. International Journal of Operations & Production Management, 6(3), 38–46. https://doi.org/10.1108/eb054765

Gaur, V., Fisher, M. L., & Raman, A. (2005). An econometric analysis of inventory turnover performance in retail services. Management Science, 51(2), 181–194. https://doi.org/10.1287/mnsc.1040.0298

Graves, S. C., & Willems, S. P. (2000). Optimizing strategic safety stock placement in supply chains. Manufacturing & Service Operations Management, 2(1), 68–83. https://doi.org/10.1287/msom.2.1.68.23267

Gruen, T. W., Corsten, D. S., & Bharadwaj, S. (2002). Retail out-of-stocks: A worldwide examination of extent, causes and consumer responses. Grocery Manufacturers of America.

Hardgrave, B. C., Aloysius, J. A., & Goyal, S. (2013). RFID-enabled visibility and retail inventory record inaccuracy: Experiments in the field. Production and Operations Management, 22(4), 843–856. https://doi.org/10.1111/poms.12010

Harris, F. W. (1990). How many parts to make at once. Operations Research, 38(6), 947–950. (Original work published 1913). https://doi.org/10.1287/opre.38.6.947

Hendricks, K. B., & Singhal, V. R. (2003). The effect of supply chain glitches on shareholder wealth. Journal of Operations Management, 21(5), 501–522. https://doi.org/10.1016/j.jom.2003.02.003

Hendricks, K. B., & Singhal, V. R. (2005). Association between supply chain glitches and operating performance. Management Science, 51(5), 695–711. https://doi.org/10.1287/mnsc.1040.0353

Hendricks, K. B., & Singhal, V. R. (2009). Demand-supply mismatches and stock market reaction: Evidence from excess inventory announcements. Manufacturing & Service Operations Management, 11(3), 509–524. https://doi.org/10.1287/msom.1080.0237

Hopp, W. J., & Spearman, M. L. (2004). To pull or not to pull: What is the question? Manufacturing & Service Operations Management, 6(2), 133–148. https://doi.org/10.1287/msom.1030.0028

Hyndman, R. J., & Koehler, A. B. (2006). Another look at measures of forecast accuracy. International Journal of Forecasting, 22(4), 679–688. https://doi.org/10.1016/j.ijforecast.2006.03.001

IFRS Foundation. (n.d.). IAS 2 Inventories. Retrieved October 24, 2026, from https://www.ifrs.org/issued-standards/list-of-standards/ias-2-inventories/

Ivanov, D. (2020). Predicting the impacts of epidemic outbreaks on global supply chains: A simulation-based analysis on the coronavirus outbreak (COVID-19/SARS-CoV-2) case. Transportation Research Part E: Logistics and Transportation Review, 136, Article 101922. https://doi.org/10.1016/j.tre.2020.101922

Kang, Y., & Gershwin, S. B. (2005). Information inaccuracy in inventory systems: Stock loss and stockout. IIE Transactions, 37(9), 843–859. https://doi.org/10.1080/07408170590969861

Kesavan, S., & Mani, V. (2013). The relationship between abnormal inventory growth and future earnings for U.S. public retailers. Manufacturing & Service Operations Management, 15(1), 6–23. https://doi.org/10.1287/msom.1120.0389

Kök, A. G., & Shang, K. H. (2007). Inspection and replenishment policies for systems with inventory record inaccuracy. Manufacturing & Service Operations Management, 9(2), 185–205. https://doi.org/10.1287/msom.1060.0136

Lee, H. L. (2002). Aligning supply chain strategies with product uncertainties. California Management Review, 44(3), 105–119. https://doi.org/10.2307/41166135

Lee, H. L., Padmanabhan, V., & Whang, S. (1997a). Information distortion in a supply chain: The bullwhip effect. Management Science, 43(4), 546–558. https://doi.org/10.1287/mnsc.43.4.546

Lee, H. L., Padmanabhan, V., & Whang, S. (1997b). The bullwhip effect in supply chains. MIT Sloan Management Review, 38(3), 93–102. https://sloanreview.mit.edu/article/the-bullwhip-effect-in-supply-chains/

Makridakis, S., Spiliotis, E., & Assimakopoulos, V. (2020). The M4 Competition: 100,000 time series and 61 forecasting methods. International Journal of Forecasting, 36(1), 54–74. https://doi.org/10.1016/j.ijforecast.2019.04.014

McKinsey Global Institute. (2020). Risk, resilience, and rebalancing in global value chains. McKinsey & Company. https://www.mckinsey.com/capabilities/operations/our-insights/risk-resilience-and-rebalancing-in-global-value-chains

Moritz, B. B., Hill, A. V., & Donohue, K. L. (2013). Individual differences in the newsvendor problem: Behavior and cognitive reflection. Journal of Operations Management, 31(1–2), 72–85. https://doi.org/10.1016/j.jom.2012.11.006

National Retail Federation. (2023). National retail security survey 2023. https://www.nrf.com/research/national-retail-security-survey-2023

Ohno, T. (1988). Toyota production system: Beyond large-scale production. Productivity Press.

Raman, A., DeHoratius, N., & Ton, Z. (2001). Execution: The missing link in retail operations. California Management Review, 43(3), 136–152. https://doi.org/10.2307/41166093

Rumyantsev, S., & Netessine, S. (2007). What can be learned from classical inventory models? A cross-industry exploratory investigation. Manufacturing & Service Operations Management, 9(4), 409–429. https://doi.org/10.1287/msom.1070.0166

Schweitzer, M. E., & Cachon, G. P. (2000). Decision bias in the newsvendor problem with a known demand distribution: Experimental evidence. Management Science, 46(3), 404–420. https://doi.org/10.1287/mnsc.46.3.404.12070

Shah, R., & Ward, P. T. (2003). Lean manufacturing: Context, practice bundles, and performance. Journal of Operations Management, 21(2), 129–149. https://doi.org/10.1016/S0272-6963(02)00108-0

Shih, W. C. (2020). Global supply chains in a post-pandemic world. Harvard Business Review, 98(5), 82–89. https://hbr.org/2020/09/global-supply-chains-in-a-post-pandemic-world

Silver, E. A., Pyke, D. F., & Thomas, D. J. (2017). Inventory and production management in supply chains (4th ed.). CRC Press.

Song, J.-S., & Zipkin, P. H. (1996). Managing inventory with the prospect of obsolescence. Operations Research, 44(1), 215–222. https://doi.org/10.1287/opre.44.1.215

Sterman, J. D. (1989). Modeling managerial behavior: Misperceptions of feedback in a dynamic decision making experiment. Management Science, 35(3), 321–339. https://doi.org/10.1287/mnsc.35.3.321

Syntetos, A. A., & Boylan, J. E. (2005). The accuracy of intermittent demand estimates. International Journal of Forecasting, 21(2), 303–314. https://doi.org/10.1016/j.ijforecast.2004.10.001

Syntetos, A. A., Boylan, J. E., & Croston, J. D. (2005). On the categorization of demand patterns. Journal of the Operational Research Society, 56(5), 495–503. https://doi.org/10.1057/palgrave.jors.2601841

Teunter, R. H., Babai, M. Z., & Syntetos, A. A. (2010). ABC classification: Service levels and inventory costs. Production and Operations Management, 19(3), 343–352. https://doi.org/10.1111/j.1937-5956.2009.01098.x

Ton, Z., & Huckman, R. S. (2008). Managing the impact of employee turnover on performance: The role of process conformance. Organization Science, 19(1), 56–68. https://doi.org/10.1287/orsc.1070.0294

Ton, Z., & Raman, A. (2010). The effect of product variety and inventory levels on retail store sales: A longitudinal study. Production and Operations Management, 19(5), 546–560. https://doi.org/10.1111/j.1937-5956.2010.01120.x

U.S. Census Bureau. (2026). Total business: Inventories to sales ratio [ISRATIO; Data set]. Retrieved October 24, 2026, from FRED, Federal Reserve Bank of St. Louis. https://fred.stlouisfed.org/series/ISRATIO

Wan, X., Evers, P. T., & Dresner, M. E. (2012). Too much of a good thing: The impact of product variety on operations and sales performance. Journal of Operations Management, 30(4), 316–324. https://doi.org/10.1016/j.jom.2011.12.002

Videos

Burke, D. (2022, March 4). How to fix broken supply chains [Video]. TED. https://www.youtube.com/watch?v=y_EjCSz1aBo

Knowledge at Wharton. (2011, July 21). Wharton professor Marshall Fisher: The new science of retailing [Video]. YouTube. https://www.youtube.com/watch?v=B4OF3L7_kco

MIT Center for Transportation & Logistics. (2023, May 15). Inventory management: How to align company strategy with the right policies? [Video]. YouTube. https://www.youtube.com/watch?v=oU6IAVwoieY

Mountz, M. (2015, July 15). What happens inside those massive warehouses? [Video]. TED. https://www.youtube.com/watch?v=9UiK4llJOXc

Stanford Graduate School of Business. (2010, August 27). Supply chain master: Prof. Hau Lee [Video]. YouTube. https://www.youtube.com/watch?v=b0dExw3es40