Safety Stock is a much used and sometimes misunderstood word among Supply Chain professionals.
- Is it the average on-hand Inventory that we should carry?
- Is it the minimum stock we should hold?
- Or the theoretical minimum?
The misunderstanding often stems from the fact that different companies have very different Inventory management practices and often use different terminologies.
This is a brief post to dispel some of the more common mistakes we make while dealing with this topic. For the benefit of wider audience and newer supply chain professionals, let us go through some of the basics first. If you are well versed with the topic, feel free to jump directly to part 2:
Understanding Safety Stock:
Imagine a supply chain without any Demand or Supply Variation. We sell (exactly) 100 units/day. We have an exact lead time of 5 days to get new inventory.
If you have 800 units this morning, will you order anything today?
You won’t.. ordering new inventory will make sense only when you are left with 500 units. To generalize, in this case you need to order just when your Inventory position (Net on-hand + on-order) falls to the demand over lead time.
In this case, the net on-hand will go to zero just when new supplies become available. So there will not be any stockout or sales loss.
Above situation simplifies and doesn’t protect against two obvious possibilities:
- What if demand is higher than forecast? (Demand variability)
- What if supply takes longer than the assumed lead time? (Supply Variability)
That’s where Safety Stock comes in picture.
Safety Stock provides a supply chain with the desired buffer against Demand and Supply Lead Times variabilities.
Think of the same scenario, but with demand and supply variability. As you can imagine, in this situation, having an inventory position at just about 500 units is risky. By the time fresh order arrives, there is a chance of us running out of stock. (if the actual demand exceeds 100 units/day)
To safeguard, we need a buffer, above picture shows the same supply chain with a safety stock of 200 units. As you can see, while Inventory may dip below safety stock buffer, it’s likely to remain above zero, safeguarding us against stockouts. In this setup, just before new replenishments arrive (the lowest on-hand position) actual stock will sometime be lower than the safety stock line (during periods of higher demand or longer lead times) and sometime higher – but on average, the on-hand stock’s minimum position should hover around that safety stock line.
Now let’s define safety stock:
Safety Stock is the average level of the net stock just before a replenishment arrives.
Key words here is “Average”. As mentioned above, On hand Stock may breach safety stock or may get replenishment before it even hits Safety Stock depending on variability.
Part 2: Common misconceptions about Safety Stocks –
Misconception1: Safety Stock (SS) is a minimum level that you should never touch.
Reality: SS is an insurance buffer designed to be consumed against uncertainty. In a healthy system, inventory will naturally dip below the SS level for part of the replenishment cycle in some cases(Mathematically speaking ~50% of replenishment cycles are likely to end with OnHand stock being lower than Safety Stock on the day when new supplies connect). In fact, If SS is never consumed, it is likely excessive or planning team is over-responding beyond the set parameters (this is true irrespective of targeted Service levels)
Actionable Takeaways: Don’t trigger a panic or RCA just because On Hand < SS.
Misconception2: Use a fixed, simple rule for all items (e.g., 30 days of coverage)
Safety Stock must be calculated dynamically based on three factors: Demand Variability, Lead Time Variability, and the desired Service Level. A high-value, highly variable “A” item needs a different calculation than a predictable “C” item.
Actionable Takeaways: Stop using fixed days of supply across all items. Embrace probabilistic methods (like using the normal distribution, Z-scores and demand variability in terms of Std deviation or RMSE) to set inventory targets that truly align with business risk and cost. These are rather simple calculations with low sensitivity around their inputs, that means even if inputs are not perfect, output will still be reasonably good.
Misconception3: Safety Stock is calculated once and remains static for months or years
Safety Stock is a dynamic variable that needs frequent recalculation. Key inputs like Demand levels, Demand Variability, Lead Times, Lead time variability change over time and hence Safety stock levels that were calculated several months ago may no longer be valid.
Actionable Takeaway: Implement a regular review cycle (monthly/quarterly) to recalculate SS. A planning software can help in more frequent refresh of safety stock levels.
Misconception4: Safety Stock Fixes Bad Forecasting or can cover other systemic issues
Safety Stock only covers random forecast error (noise), not systematic bias (signal). Most companies fail to apply proper Safety Stock formula to count for a known, persistent bias (more on that in a separate post someday) and usually end up using regular Safety Stock formula on a bias inflated (or deflated) forecasting data. Having a bias, and not handling it properly will usually lead us to higher Safety stock. Inflating SS to cover a consistently high or low forecast bias is expensive and masks the real problem—a flawed forecasting model or process. Similiarly having a wrong BOM, or inaccurate lead time are kind of errors that can’t be hidden/covered by safety stock in most cases. Having high safety stock can mask them but this masking will often lead to far mor serious issues in future (as systemic errors are not uncovered in time)
Actionable Takeaway: Before blindly increasing SS, address the root cause of large errors.
Misconception5: Looking at Safety Stock in Silo as the only controllable factor to ensure Service levels.
Safety Stock is one of several levers for ensuring Supply Chain agility. Other levers involved are: holding Finished Goods (FG) vs. Component/Raw Material inventory, Supplier Dual-Sourcing, Strategic Capacity Planning, or shortening the Order Lead Times. These other levers are oftenmore cost-effective ways to improve resilience than just simply increasing SS at finished goods end.
Actionable Takeaway: Safety stock is the last line of defense. This buffer will cover for demand and supply variabilities as seen in past or predicted for future, but it doesn’t cover for gaps in supply chain design, and if often very costly to maintain. Invest in improving upstream stability (supplier performance, lead time reduction) to ultimately reduce the need for excessive safety stock.
If you are a supply chain professional, I’d love to hear about your views – What other misconception have you seen around safety stocks.