What Is the Z-Score Service Factor?
The Z-score service factor is the number of standard deviations of safety stock you carry above your average lead time demand. It is the most important variable in the safety stock formula, because it is the only one you control directly. Demand variability and lead time are facts about your supply chain. The Z-score service factor is a policy decision: how much risk of a stockout am I willing to accept?
Every Z-score corresponds to a cumulative probability under the standard normal distribution curve. A Z-score of 1.65 covers 95% of demand outcomes, meaning that across many replenishment cycles, you will avoid running out 95% of the time. A Z-score of 2.33 pushes that to 99%, but at the cost of carrying significantly more inventory.
The Z-score service factor is sometimes just called the "service factor" in inventory textbooks. They are the same thing. When you read "service factor of 1.65" in an SAP manual or APICS exam, it means Z = 1.65, which means a 95% target service level.
Z-Score Service Factor Formula and Lookup Table
The Z-score service factor enters the safety stock formula as the multiplier on demand variability:
Safety Stock = Z × σ_dLTWhere Z is the Z-score service factor and σ_dLT is the standard deviation of demand during lead time. Common values from the standard normal table:
| Service Level | Z-Score | Excel Formula | Use Case |
|---|---|---|---|
| 80% | 0.84 | =NORM.S.INV(0.80) | Low-margin clearance SKUs |
| 90% | 1.28 | =NORM.S.INV(0.90) | C-tier SKUs with low contribution |
| 95% | 1.65 | =NORM.S.INV(0.95) | Default for mid-tier FBA SKUs |
| 97.5% | 1.96 | =NORM.S.INV(0.975) | High-margin or hero ASINs |
| 99% | 2.33 | =NORM.S.INV(0.99) | Critical SKUs, premium brands |
| 99.5% | 2.58 | =NORM.S.INV(0.995) | Subscribe & Save anchor SKUs |
Notice how the marginal cost rises. Moving from 95% to 99% increases the Z-score service factor from 1.65 to 2.33, a 41% jump in safety stock. Moving from 99% to 99.5% adds another 11%. Each percentage point of service level gets more expensive.
Worked Example: Choosing a Z-Score for Two SKUs
You have two FBA SKUs. Both have the same average daily demand and lead time, but very different economics.
SKU A: Hero kitchen tool. ASP $42, contribution margin $14/unit. Subscribe & Save ratio 35%. Stockout = lost subscription revenue plus BSR damage.
SKU B: Seasonal accessory. ASP $19, contribution margin $3/unit. Sells through unevenly. Stockout = a few lost sales, no lasting damage.
Both have σ_dLT = 80 units (standard deviation of lead time demand).
SKU A choice: Z = 2.33 (99% service level). Safety Stock = 2.33 × 80 = 186 units.
SKU B choice: Z = 1.28 (90% service level). Safety Stock = 1.28 × 80 = 102 units.
SKU A carries 84 more units of buffer than SKU B. At a $9 landed cost per unit, that is $756 more cash tied up. But because SKU A's contribution margin is $14 versus $3, the extra buffer pays for itself after just 6 months of stockouts avoided. SKU B does not justify the extra buffer.
This is the practical use of the Z-score service factor: matching buffer to economics, not applying one number to every SKU.
FBA-Specific Context for Z-Score Selection
Generic supply chain advice says "use Z = 1.65 for everything." For FBA, that is too crude. Three Amazon-specific factors should adjust your Z-score service factor up or down:
1. Subscribe & Save penalty. If a SKU has a meaningful Subscribe & Save base, a stockout cancels recurring orders and Amazon may reduce your S&S allocation for weeks. Use Z = 1.96 to 2.33 minimum on these SKUs.
2. Storage limit risk. If your FBA storage limits are tight, every unit of safety stock crowds out other SKUs. A high Z-score for one SKU can starve another. Balance Z-scores across the catalog so your total buffer fits within your shipment plan.
3. Replenishment lead time mix. If you replenish via Amazon Warehousing and Distribution (AWD) with shorter cycles, you can run lower Z-scores because you can react faster. Direct-to-FBA from Asia at 60+ days warrants a higher Z to absorb supply risk.
Once you've picked the right Z, plug it into our free safety stock calculator to see the buffer in units.
Common Mistakes
1. Using one Z-score for the whole catalog. Applying Z = 1.65 to all SKUs means your hero ASINs are under-protected and your D-tier SKUs are over-stocked. The whole point of the Z-score service factor is to differentiate by SKU economics.
2. Confusing the Z-score with the service level. A Z-score of 2.33 is not a 2.33% service level. It is the standard deviation count that produces a 99% probability under the normal curve. Mixing these up leads to massive over-buffering.
3. Ignoring the marginal cost curve. Going from 95% to 99% sounds like only 4 percentage points, but it is a 41% jump in safety stock cost. Always check the cash impact of pushing the Z-score service factor higher before committing.
Related Terms
Frequently Asked Questions
What Z-score service factor should I use for Amazon FBA?
Most FBA sellers use Z = 1.65 for a 95% service level on mid-tier SKUs and Z = 1.28 for 90% on slower movers. Hero ASINs with high contribution margin often run at Z = 1.96 (97.5%) or higher. Match your Z-score to the cost of stocking out, not just round numbers.
How is Z-score different from service level?
Service level is a percentage (the probability of not stocking out). Z-score is the standard normal value that corresponds to that probability. They are two ways of expressing the same target. The Z-score service factor is what plugs into the safety stock math; service level is what you communicate to leadership.
Why does each percentage point of service level cost more buffer?
The normal distribution has fat shoulders and thin tails. Going from 95% (Z = 1.65) to 99% (Z = 2.33) requires 41% more safety stock. Going from 99% to 99.9% requires another 32%. Each marginal point of service level costs disproportionately more cash.
Can I use a Z-score below 1?
Yes. A Z = 0.84 corresponds to an 80% service level. This is appropriate for low-margin clearance SKUs where the cost of holding extra units exceeds the cost of occasional stockouts. A Z below zero means you target less than 50% service level and are accepting frequent stockouts.
Is the Z-score the same in Excel as in my safety stock formula?
Yes. In Excel, NORM.S.INV(0.95) returns 1.6449, which is the Z-score for a 95% service level. NORM.S.INV(0.975) returns 1.96. These are the exact values you plug into Safety Stock = Z × σ_dLT.