Lead Time and Demand Variability: The Two Numbers That Drive Every Reorder
Lead time and demand variability are the two inputs that determine every reorder point and safety stock level.
Every reorder decision comes down to two numbers: how long it takes stock to arrive (lead time), and how unpredictably customers consume it (demand variability). Get those two numbers right and your reorder point almost calculates itself. Get them wrong, and you are either sitting on cash-eating overstock or watching your best-seller go out of stock while the replacement order sits on a container ship.
Key takeaways
- Reorder Point = (Average Daily Sales x Average Lead Time) + Safety Stock. Two inputs drive everything.
- Demand variability and lead time variability are separate problems that compound each other when both are high.
- A gut-feel reorder based on "it feels low" ignores the math and almost always fires too late or too early.
- Safety stock is not a buffer you set once. It should be recalculated whenever your supplier's lead time changes or a SKU's sales pattern shifts.
- In 2026, supply chains are structurally more volatile: geopolitical disruptions now occur roughly every 3.7 years on average, making static reorder points riskier than ever.
What lead time actually means for your inventory
Lead time is the number of calendar days between placing a purchase order and receiving usable stock on your shelf (or in your warehouse). It sounds simple. It is not.
Most merchants track a single "supplier lead time" number, often pulled from memory or the supplier's sales pitch. The more useful figure is lead time variability: the standard deviation across your last 10 to 20 orders from that supplier. A supplier who delivers in "about 14 days" might actually range from 9 to 22 days. That 13-day spread is the number that determines how much safety stock you truly need.
Practically, lead time has three components worth separating:
- Order processing time: how long the supplier takes to acknowledge and pick the order.
- Transit time: carrier and customs dwell time (this is where most variability hides).
- Receiving time: your own inbound check-in and put-away lag before stock is sellable.
All three add up. If you only track the carrier tracking date and ignore the receiving lag, your reorder point is consistently understated.
What demand variability means, and why average daily sales mislead you
Demand variability measures how much your actual daily (or weekly) sales deviate from the average. A SKU that sells exactly 10 units every single day has zero demand variability. A SKU that sells 0 on Monday and 40 on Saturday has high variability even if the weekly average is still 70 units.
The classic mistake is using a simple average daily sales figure in the reorder point formula. That average looks stable on a spreadsheet, but it masks spikes that will drain your stock in two days instead of seven. The correct metric is the standard deviation of daily demand over a rolling 30 to 90 day window, updated regularly.
High demand variability means you need more safety stock. High lead time variability means you need even more. When both are high simultaneously, the buffer required grows faster than most merchants expect.
The reorder point formula, written out plainly
The full reorder point (ROP) formula that accounts for both sources of variability is:
ROP = (Average Daily Sales x Average Lead Time) + Safety Stock
Safety Stock = Z x sqrt(Average Lead Time x σ_demand² + Average Daily Sales² x σ_lead_time²)Where:
- Z = service-level z-score (1.65 for 95% in-stock, 2.05 for 98%)
- σ_demand = standard deviation of daily demand
- σ_lead_time = standard deviation of lead time in days
The key insight is that both variabilities are inside a square root together. That means reducing either one lowers your required safety stock, freeing up cash. A merchant who negotiates a more consistent lead time from their supplier (reducing σ_lead_time) can carry less safety stock without lowering their service level at all. That is a concrete financial argument for supplier performance tracking.
To make this concrete: imagine a SKU selling an average of 20 units per day, with an average lead time of 14 days and no variability in either. The base ROP is 280 units (20 x 14). Now add a demand standard deviation of 5 units/day and a lead time standard deviation of 3 days, at a 95% service level (Z = 1.65). Safety stock comes out to roughly 115 units, raising the ROP to 395 units. Ignore variability and you will stock out 5% of the time on that SKU alone.
When gut feel breaks down (and the cost of getting it wrong)
Gut feel works when you have fewer than 20 SKUs, a single supplier, and consistent demand. Once you cross into dozens or hundreds of SKUs, gut feel introduces a predictable failure mode: you reorder the SKUs you remember worrying about, not the ones that are statistically closest to a stockout.
The cost of a single stockout on a best-selling SKU is not just the missed sale. It is the lost-revenue cascade: customers who would have added complementary items, customers who switch to a competitor and do not come back, and the expediting cost of an emergency air freight order to recover. A SKU that generates $500/day in margin going out of stock for 10 days costs $5,000 in direct margin, before any of those compounding effects.
Conversely, setting reorder points too high because you are anxious about stockouts ties up working capital in slow-moving stock. That cash is unavailable for marketing, product development, or the next purchase order on a faster-moving SKU.
How supply chain volatility in 2026 changes your calculation
Static reorder points were always a simplification. In 2026 they are actively dangerous. According to a 2025 World Economic Forum report, major supply chain disruptions now occur approximately every 3.7 years, and each one can take two to three years to fully recover from. The practical implication for ecommerce merchants: lead time data from two years ago is often stale, and a supplier's quoted lead time today may not reflect their actual performance under current freight conditions.
Earlier this year, global e-commerce brands began shifting from lean "just-in-time" models toward more flexible, resilient supply networks built specifically for volatility. That shift means actively tracking your actual received lead time on every PO, not just relying on a number you entered into a spreadsheet in 2023. It also means recalculating your ROP more frequently than the once-a-month cadence that was once considered standard.
The merchants best insulated from this volatility are the ones who track σ_lead_time per supplier as a live metric, not a one-time estimate.
A practical 4-step process to set your reorder points correctly
- Pull 60 to 90 days of actual sales data per SKU. Calculate the mean daily demand and the standard deviation. Do not use a single average across a long period that includes outlier events (a viral post, a flash sale) without flagging those days.
- Pull your last 10 to 20 completed purchase orders per supplier. Calculate the mean lead time and the standard deviation from PO-sent date to stock-available date.
- Decide your service level per SKU. Not everything needs 98%. A high-margin, high-velocity SKU with no substitute justifies a higher Z-score (more safety stock). A slow mover with an easy substitute can run at 90%.
- Plug both standard deviations into the full ROP formula above. Set a calendar reminder to recalculate whenever a supplier changes their terms, a SKU's velocity shifts by more than 20%, or you enter a seasonal period.
For merchants managing hundreds of SKUs, this process does not scale manually. That is where tooling earns its cost.
Stockcast: Inventory Forecast runs exactly this logic against your Shopify sales history and supplier lead times every day, surfaces which SKUs are closest to their reorder point ranked by urgency, and shows you the underlying math rather than just a colored warning badge. If you want to see how your current reorder points compare to what the formula actually suggests, the free plan covers up to 25 SKUs with no trial limit.
Reorder approach comparison: which method fits your situation
| Approach | When to use | Key trade-off |
|---|---|---|
| Fixed reorder point (manual) | Under 20 SKUs, stable demand, one supplier | Fast to set up, breaks down with variability or scale |
| Average-based ROP (no variability) | Stable, predictable SKUs only | Underestimates safety stock; stockouts during spikes |
| Full variability-adjusted ROP | Any SKU with seasonal or erratic demand | Requires tracking σ_demand and σ_lead_time per SKU |
| Periodic review (time-based) | Commodities with fixed supplier windows | Simple, but misses intra-period stockout risk |
| Automated daily recalculation | 50+ SKUs, multiple suppliers, or locations | Most accurate; requires software or a maintained model |
FAQ
What is the difference between lead time variability and demand variability?
Lead time variability measures how inconsistent your supplier's delivery time is across orders. Demand variability measures how much your customers' daily or weekly purchase rate fluctuates. Both increase the safety stock you need to avoid a stockout. When both are high at the same time, the required buffer grows substantially because the two risks can compound.
How often should I recalculate my reorder points?
For fast-moving or high-value SKUs, recalculate monthly at a minimum, and immediately after any supplier lead time change or a significant shift in sales velocity. For stable, slow-moving SKUs, quarterly is usually sufficient. The key trigger is change: a new supplier, a new sales channel, or entering a seasonal period all justify an immediate recalculation.
Can I just add a flat percentage buffer to my reorder point instead of calculating safety stock properly?
You can, but a flat percentage buffer does not scale correctly across SKUs with different variability profiles. A SKU with very consistent demand needs far less buffer than a SKU with erratic demand, and applying the same flat percentage to both will leave one understocked and one overstocked. The formula-based approach sizes the buffer to the actual statistical risk, which is more accurate and uses less working capital overall.
Frequently asked questions
What is the difference between lead time variability and demand variability?
Lead time variability measures how inconsistent your supplier's delivery time is across orders. Demand variability measures how much your customers' daily or weekly purchase rate fluctuates. Both increase the safety stock you need to avoid a stockout. When both are high at the same time, the required buffer grows substantially because the two risks can compound.
How often should I recalculate my reorder points?
For fast-moving or high-value SKUs, recalculate monthly at a minimum, and immediately after any supplier lead time change or a significant shift in sales velocity. For stable, slow-moving SKUs, quarterly is usually sufficient. The key trigger is change: a new supplier, a new sales channel, or entering a seasonal period all justify an immediate recalculation.
Can I just add a flat percentage buffer to my reorder point instead of calculating safety stock properly?
A flat percentage buffer does not scale correctly across SKUs with different variability profiles. A SKU with very consistent demand needs far less buffer than a SKU with erratic demand, and applying the same flat percentage to both will leave one understocked and one overstocked. The formula-based approach sizes the buffer to the actual statistical risk, which is more accurate and uses less working capital overall.