Two Problems That Look Like Opposites
Walk into most distribution or manufacturing businesses and you will hear two complaints from two different people on the same morning.
Sales will tell you the fast-moving lines are out of stock again and a customer has gone elsewhere. Finance will tell you there is too much money sitting in the warehouse and some of it has been there for years.
These sound like contradictory problems and they are the same problem. Both are symptoms of replenishment decisions being made without a current view of demand and lead time: some items get ordered too late, others get ordered too often, and nobody is systematically deciding which is which.
What makes it persistent is that it never becomes anyone’s project. A stockout is handled by an expedited order. Overstock is handled by not thinking about it. Neither triggers a review of how reorder points were set, which in most businesses was some years ago, by someone who has since changed role.
This article covers how reordering rules work, how to set a reorder point that reflects actual demand and lead time rather than a guess, what safety stock is genuinely for, why segmenting your items before setting anything is the step that makes the rest work, and how to build a review cycle that survives. It is written for supply chain managers, operations directors, purchasing managers and the finance leaders who fund the stock.
The Business Problem: Replenishment Decided by Memory
- Reorder points were set once. Usually at implementation, often from a round number, occasionally from a formula nobody kept. Demand has moved since. The numbers have not.
- Lead times are assumed rather than measured. The purchasing team believes a supplier takes two weeks because that is what the supplier said. What the supplier actually delivers, measured across the last twenty orders, is a different and more variable number.
- Replenishment runs on individual recall. A buyer who knows the business notices when something is getting low. The system is a record of what happened, not a trigger for what should happen next.
- Safety stock is a blanket percentage. Add twenty per cent to everything. This over-protects the predictable items and under-protects the volatile ones, which is precisely backwards.
- Nobody distinguishes between item types. The item ordered weekly in large volumes and the item ordered twice a year are managed by the same rule, which cannot be right for both.
- Slow-moving and obsolete stock is never formally identified. It sits in the warehouse, counted in inventory value, until a write-off eventually forces the conversation.
- Minimum order quantities and pack sizes distort everything. A supplier’s minimum order is a box of five hundred for an item you sell forty of a year. That is a decision, and it is usually made by whoever is placing the order rather than by policy.
What Getting Replenishment Wrong Costs
- Lost sales you never see. A stockout on a fast-moving line does not usually produce a complaint. It produces a customer who buys elsewhere this time, and sometimes next time too. It is the least visible cost in this entire list and frequently the largest.
- Working capital in the wrong places. Capital tied up in slow movers is capital unavailable for the items that actually turn. The total inventory figure can look reasonable while being composed almost entirely of the wrong things.
- Expediting costs that nobody attributes. Air freight instead of sea, premium supplier pricing for a rush order, a production run interrupted. These are usually absorbed into general costs rather than attributed to the replenishment failure that caused them.
- Obsolescence. Stock that was ordered in a quantity that made sense at the time, for demand that then changed. Every month it stays, it becomes less likely to sell at full value.
- Storage and handling consumed by the wrong stock. Warehouse space is finite. Slow movers occupy locations that fast movers would turn through many times a year.
- Purchasing time spent on the wrong things. A buyer manually watching stock levels is a buyer not negotiating with suppliers or managing exceptions.
How Reordering Rules Work
The underlying mechanism is simple, and its simplicity is why it is often set up badly.
A reordering rule associates a product and a location with a minimum quantity and a maximum quantity. When available stock falls below the minimum, replenishment is proposed up to the maximum. Depending on how the product is configured, that proposal becomes a purchase order to a supplier, a manufacturing order, or an internal transfer from another location.
Three points determine whether this works in practice.
- Available stock is not the same as stock on hand. A correctly configured rule considers what is already on order and what is already committed to customers. A business that triggers on physical quantity alone will order again for stock that is already arriving.
- The rule operates per location. In a multi-warehouse business this is a design decision, not a detail: are you replenishing each location independently from suppliers, or centrally into a main warehouse and then distributing? Both are valid and they produce very different rules.
- The maximum determines order size, and therefore order frequency. The gap between minimum and maximum is the quantity you will buy each time. Set it too small and you place frequent small orders with high handling cost; set it too large and you buy months of cover for an item whose demand may change.
The usual caveat: capability and configuration options vary by Odoo version and edition. Confirm the specifics for your own instance with your partner rather than assuming any particular behaviour.
Setting a Reorder Point That Reflects Reality
The reorder point answers one question: how much stock do I need on hand now to survive until the replenishment arrives?
That makes it a function of two things (how fast the item sells, and how long it takes to arrive) plus a buffer for the fact that neither is constant.
- Start with measured demand, not remembered demand. Pull actual issue or sales history for a period long enough to be meaningful and recent enough to be relevant. For most businesses, twelve months of history reviewed for the last three months’ trend is a reasonable basis. Watch for seasonality, for one-off large orders that distort the average, and for items whose demand has structurally changed rather than simply varied.
- Measure lead time from your own data, not from the supplier’s claim. The gap between order date and receipt date, across your recent orders, for that supplier and ideally that item. Two numbers matter: the average and the spread. A supplier averaging fourteen days with a range of twelve to sixteen is a completely different planning problem from one averaging fourteen days with a range of seven to thirty-five.
- The core calculation is demand during lead time, plus a buffer. Average demand per day multiplied by average lead time in days gives you what you will consume while waiting. The buffer, safety stock, covers the variability in both figures. That is the whole model, and it is enough for the great majority of items.
- Add the handling realities. Supplier minimum order quantities, pack sizes, and your own receiving capacity all constrain what you can actually order. These belong in the maximum quantity rather than in the reorder point, and where a minimum order quantity represents years of demand, the right answer is usually to question the item or the supplier rather than to absorb it.
- Do not model what you cannot maintain. Sophisticated statistical models are available and they are frequently the wrong choice, because a model nobody can explain is a model nobody will update. A simple calculation, reviewed quarterly, outperforms a sophisticated one applied once and abandoned.
Safety Stock: What It Is Actually For
Safety stock exists to cover variability, not to cover demand. That distinction is the one most businesses get wrong, and getting it wrong is what produces a blanket percentage applied to everything.
- It protects against two uncertainties. Demand being higher than average during the lead time, and the lead time being longer than average. Items where both are stable need very little. Items where either is volatile need more.
- It should therefore differ by item, and the differences should be large. A product with steady weekly demand from a reliable local supplier might need days of cover. A product with lumpy demand from an overseas supplier whose delivery performance varies by weeks needs substantially more. Applying the same percentage to both guarantees you are wrong about at least one.
- The right level is a commercial decision, not a mathematical one. The real question is what a stockout costs on this particular item. For a component that halts a production line, the answer is high and the safety stock should reflect it. For an item a customer will happily wait three days for, the answer is low. Someone has to make that judgement per item group, and it is a conversation between operations, sales and finance rather than a calculation.
- Review it when the inputs change, not on a fixed timetable alone. A supplier change, a new source country, a significant demand shift, or a change in customer service commitment should all trigger a review of the items affected.
Segment Before You Set Anything
This is the step that makes everything else achievable, and it is the one most often skipped because it feels like analysis rather than action.
You cannot manage several thousand items with equal attention, and you should not try. Segment them, then apply different levels of effort to different segments.
- By value and volume. The familiar approach: a small proportion of items usually accounts for the majority of consumption value. Those items deserve measured demand, measured lead time, individually considered safety stock and frequent review. The long tail does not.
- By demand pattern. Steady, seasonal, lumpy, or declining. Each needs different treatment, and lumpy demand in particular is where reorder points based on averages fail most badly: an item ordered four times a year in large quantities is not well served by an average daily demand figure.
- By criticality. What happens when this item is unavailable? A line-stopping component and a replaceable consumable deserve different buffers regardless of their value.
- By supplier reliability. Items from suppliers with variable delivery performance need more protection than identical items from reliable ones.
What to do with the result: manage the top segment actively, with measured inputs and quarterly review. Manage the middle with simpler rules and annual review. For the long tail, use simple rules, accept occasional stockouts as cheaper than the effort of preventing them, and spend the saved attention on the items that matter.
And review the tail for a different purpose. The long tail is also where obsolete stock accumulates. A scheduled review that asks “should we still be stocking this at all?” belongs in the same exercise.
The Review Cycle Nobody Schedules
Replenishment settings are not configuration; they are a living policy. The failure mode in every business that has done this work once is that they never do it again.
- Make it a scheduled task with a named owner. Quarterly for the top segment, annually for the rest. In a calendar, with a person’s name against it.
- Review exceptions continuously, not just settings periodically. A short report of items that stocked out, items where the rule triggered unusually often, and items with no movement in a defined period. That report is what tells you which settings are wrong without reviewing all of them.
- Re-measure lead times rather than assuming them. Supplier performance drifts, and it drifts in both directions. The measurement is a report, not a project.
- Report the outcomes, not the activity. Stockout incidents on the top segment, inventory value by segment, stock turns, and the value of items with no movement. Four numbers, reported monthly, will keep this alive. Reporting how many rules were reviewed will not.
- Treat a stockout as a question. Not “how do we get stock in quickly” (that is the immediate response) but “was the reorder point wrong, was the lead time longer than assumed, or was demand genuinely abnormal?” Each answer leads somewhere different, and asking the question is what converts a recurring problem into a solved one.
Replenishment Approaches Compared
| Approach | How it decides | Effort | Best used for | Main failure mode |
|---|---|---|---|---|
| Buyer judgement | A person notices stock is low | High, and concentrated in one person | Very small item counts | Key-person dependency; invisible stockouts |
| Fixed reorder point, set once | Static minimum and maximum | Low | Nothing, beyond the first year | Settings drift out of line with demand and lead time |
| Reorder point, reviewed on a cycle | Static rule, periodically re-measured | Moderate, and predictable | The majority of items in most businesses | Only fails if the review stops happening |
| Segmented rules with differentiated review | Different effort per segment | Moderate, correctly targeted | Businesses with wide item ranges | Requires maintained segmentation |
| Forecast-driven planning | Projected demand over a horizon | High; needs data quality and skill | Seasonal, promotional or long-lead-time items | A forecast nobody owns is worse than a reorder point |
For most mid-market distributors and manufacturers, segmented reorder points with a real review cycle outperform both the buyer-judgement model they are replacing and the forecasting model they imagine they need.
Best Practices
- Segment the item range before setting any rule. Different items deserve different attention, and pretending otherwise is why the work never gets maintained.
- Measure lead times from your own receipt history, and record the spread as well as the average.
- Base demand on measured history, with one-off large orders excluded and seasonality identified.
- Set safety stock from variability and criticality, item group by item group, never as a blanket percentage.
- Put minimum order quantities and pack sizes into the maximum, not into the reorder point.
- Question any item whose supplier minimum represents more than a year of demand. That is a sourcing conversation, not a stock setting.
- Schedule the review with a named owner: quarterly for the top segment, annually for the rest.
- Run a monthly exception report: stockouts, unusually frequent triggers, and zero-movement items.
- Ask the diagnostic question after every stockout: wrong reorder point, longer lead time, or abnormal demand?
- Report four outcome numbers monthly: stockout incidents on the top segment, inventory value by segment, stock turns, and value of non-moving stock.
Common Mistakes
- A blanket safety stock percentage. Over-protects the predictable and under-protects the volatile, exactly backwards.
- Using the supplier’s stated lead time. Measure what they actually deliver, including the variability.
- Setting rules on physical stock rather than available stock. Produces duplicate ordering for stock already on its way.
- Treating all items the same. Guarantees that effort is spread too thin to help the items that matter.
- Averaging lumpy demand. An item bought four times a year is badly served by an average daily figure.
- Absorbing supplier minimum order quantities without question. Years of cover arrives silently and becomes obsolete stock.
- Setting it up once and never reviewing. The single most common failure, and the reason this work gets redone every few years.
- Ignoring the long tail entirely. It is where obsolescence accumulates, and it needs a different question: not a better reorder point, but whether to stock the item at all.
- Building a forecast model nobody owns. More sophisticated and less effective than a simple rule that gets reviewed.
An Illustrative Scenario: An Electrical Wholesaler
The following is a composite illustration drawn from patterns common to wholesale distribution. It is not an account of a specific Techvaria client, and no figures below come from any real engagement.
Consider an electrical wholesaler carrying several thousand stock items across two branches, supplying contractors and smaller retailers.
The Presenting Complaint
The branch managers said they were constantly out of the lines that sold most. The finance director said inventory value had grown faster than revenue for two years running. Both were right.
What the Segmentation Showed
Sorting the range by consumption value produced the expected shape: a minority of items accounting for the great majority of value. More usefully, sorting by movement showed a substantial number of items that had not moved at all in over a year, and a smaller number that had stocked out repeatedly.
The repeated stockouts were concentrated in the top value segment. The non-moving items were almost entirely in the tail. The business had been applying the same replenishment approach to both.
What the Lead Time Measurement Showed
Purchasing had been planning on supplier-quoted lead times. Measured against actual receipts, one significant supplier was averaging close to their quoted figure but with a very wide spread, occasionally far quicker, occasionally far slower. The reorder points on that supplier’s items had been set for the average, which meant they were adequate most of the time and inadequate precisely when it mattered.
What Was Implemented
Segmentation first, into three tiers by consumption value, cross-cut by criticality. Lead times re-measured per supplier from receipt history, with the spread recorded alongside the average.
Reorder points for the top tier were rebuilt from measured demand and measured lead time, with safety stock set per item group from variability and criticality rather than as a percentage. The middle tier got simpler rules. The tail kept simple rules and gained something more useful: a scheduled review asking whether each item should be stocked at all.
Several supplier minimum order quantities were challenged. Two items were moved to a different supplier; three were discontinued, with the remaining stock cleared.
A quarterly review was scheduled for the top tier with a named owner, and a monthly exception report covering stockouts, over-frequent triggers and zero-movement items was added to the operations meeting.
After Two Quarters
Stockouts on the top tier reduced noticeably, not because safety stock increased everywhere, but because it increased where variability justified it and decreased where it did not. Inventory value moved less than expected in total, but its composition shifted: less capital in the tail, more in items that turned.
The finding the operations director considered most valuable had nothing to do with reorder points. The zero-movement review surfaced a body of stock nobody had looked at in years, and dealing with it released warehouse space that had been quietly constraining the fast-moving lines.
Industry Use Cases
- Trading and distribution. The core case: wide ranges, thin margins, and stockouts that lose sales to a competitor the same day. See trading and distribution.
- Manufacturing. Component availability where criticality dominates value, because a cheap part that halts a line costs more than an expensive one that does not. See manufacturing solutions.
- Automotive parts and accessories. Very large ranges with pronounced long tails, where the stock-it-or-not question matters as much as the reorder point. See automobile and EV solutions.
- Healthcare and medical supplies. Criticality and expiry together, where overstock has a shelf-life consequence as well as a capital one. See healthcare solutions.
- E-commerce and retail. Demand volatility and promotional spikes that make average-based reorder points unreliable without seasonality handling. See e-commerce solutions.
Frequently Asked Questions
Average demand per day multiplied by average lead time in days, plus safety stock to cover variability in both. That is sufficient for the great majority of items. The accuracy comes from measuring demand and lead time from your own data rather than from assumptions, not from using a more complex formula.
It should differ by item, driven by how variable the demand and lead time are and by what a stockout costs on that particular item. A blanket percentage applied across the range over-protects stable items and under-protects volatile ones. Set it by item group, with operations, sales and finance agreeing what a stockout is worth avoiding.
For most mid-market businesses, no, not as a first step. Segmented reorder points with a genuine review cycle outperform a forecast model that nobody owns or maintains. Forecasting earns its place for seasonal, promotional or very long lead time items, and it is better added to a working replenishment process than used to replace a broken one.
Quarterly for your highest-value segment, annually for the rest, with a named owner and a calendar entry. Alongside that, run a monthly exception report so you learn which settings are wrong without having to review all of them.
Treat it as a sourcing question rather than a stock setting. Where a minimum represents more than a year of demand, the realistic options are an alternative supplier, a different pack configuration, consolidating with other orders, or discontinuing the item. Absorbing it quietly is how obsolete stock is created.
Reordering rules operate per product and location, so multi-location replenishment is a design decision: whether each location replenishes independently from suppliers, or centrally into a main warehouse with onward distribution. Both are workable and they produce different rule structures, so decide it deliberately. Confirm the specific configuration options for your version and edition with your partner.
Do not try to optimise them. Use simple rules, accept occasional stockouts as cheaper than the attention required to prevent them, and run a scheduled review asking a different question: whether the item should be stocked at all. That review is usually where the most capital is released.
Conclusion
Being out of stock on the lines that sell and overstocked on the lines that do not is one problem, not two, and it is caused by applying the same replenishment approach to items that deserve very different attention.
The fix has three parts and none of them is sophisticated. Segment the range so effort goes where it is worth spending. Set reorder points from measured demand and measured lead time, with safety stock sized to variability and criticality rather than applied as a percentage. And schedule a review with a named owner, because the single most reliable failure in this area is doing the work once and never returning to it.
The last part matters most. Businesses that implement careful reorder points and never review them are back where they started within about two years, which is why this work so often gets redone rather than maintained. A simple calculation reviewed quarterly will beat a sophisticated one applied once, every time.
If you are starting, start with the measurement: actual lead times from your own receipts, and a movement analysis of the full range. Both are reports rather than projects, and between them they will tell you how big your problem is and exactly where it sits.
Find Out Which of Your Reorder Points Are Wrong
Techvaria is an official Odoo Silver Partner with more than 350 implementations delivered since 2016, from offices in Bangalore, Ahmedabad and Dubai.
A replenishment review works from your own data and produces four things: your range segmented by consumption value, movement and criticality; measured supplier lead times with their spread, compared against what you are currently planning on; a list of items whose reorder points are demonstrably wrong, ranked by what that is costing; and a review cycle with owners, so the work stays done.
It usually also produces an uncomfortable list of items that have not moved in a year, which is where the recoverable capital tends to be.
Talk to our Odoo consultants, or see how we approach implementation.
Find Out Which of Your Reorder Points Are Wrong
Mustufa Rahi is an Odoo Certified Functional Consultant and ERP expert at Techvaria with 15+ years of experience in implementation, automation, and business process optimization, helping organizations scale efficiently.