There's something particularly frustrating about running out of stock.
Especially when you thought you had enough.
A customer places an order. The stock isn't there. Somebody raises the alarm, and suddenly everyone's asking the same question.
How did we run out?
It's a familiar situation in supply chain.
And the temptation is to start looking for somebody, or something, to blame.
Was the forecast wrong?
Did the supplier let us down?
Did somebody forget to place an order?
Perhaps.
But sometimes the explanation is sitting quietly in the system, exactly where it's been for months.
The inventory settings.
And one number in particular.
Safety stock.
A number we learn to trust
I was reminded of this recently when a stockout prompted me to take a closer look at a product's inventory settings.
Nothing particularly unusual about that.
Products run out of stock. Demand changes. Supply doesn't always behave as expected.
But something caught my attention.
The safety-stock setting looked low compared with the recent sales activity.
So I looked back over the previous months.
Demand wasn't entirely predictable, but there was enough in the history to make me question whether we were holding sufficient protection.
I didn't spend hours investigating it.
There were customers waiting for their orders.
I explained what I'd found, suggested that the safety stock needed reviewing and got the process moving to replenish the product.
The immediate problem was getting stock back into the business.
The longer-term question was whether our replenishment settings still made sense.
And it got me thinking about how much confidence we sometimes place in the numbers sitting inside our planning systems.
Because once a safety-stock figure has been calculated, approved and entered into an ERP system, it's remarkably easy for it to become something we simply accept.
Thirty units.
A hundred units.
Five hundred units.
Whatever the number happens to be.
It's there.
The system uses it.
And we assume it's doing its job.
But when was it last reviewed?
And perhaps more importantly, what assumptions were used to calculate it in the first place?
What is safety stock actually protecting us from?
Let's go back to the fundamentals.
Safety stock is additional inventory held to protect against uncertainty.
That uncertainty might come from customer demand.
Perhaps customers order more than expected, or their ordering patterns are difficult to predict.
It might come from supply.
A supplier quotes a four-week lead time, but sometimes delivers in five.
Or six.
Or perhaps both demand and replenishment are unpredictable.
Safety stock provides a buffer against that uncertainty.
But here's an important distinction.
Safety stock isn't a guarantee that you'll never run out.
It's a level of protection designed around a particular set of assumptions and an acceptable level of risk.
And that risk matters.
A business targeting a 95% cycle service level is making a different inventory decision from one targeting 99%.
The second target generally requires more safety stock, assuming the same demand and supply conditions.
But more stock comes at a cost.
Working capital.
Storage.
Insurance.
Obsolescence.
Potentially even disposal.
So the objective isn't to eliminate every possible stockout at any cost.
It's to understand the risk, decide what level of protection is appropriate and hold the inventory needed to support that decision.
Which sounds straightforward.
Until the assumptions start changing.
Safety stock isn't the whole calculation
There's another important concept that often gets mixed up with safety stock.
The reorder point.
In a basic continuous-review replenishment system, the reorder point is the inventory position at which we trigger another order.
And it needs to cover two things.
The stock we expect to consume while waiting for replenishment.
And the additional stock we hold to protect against uncertainty.
In its simplest form:
Reorder Point = Expected Demand During Lead Time + Safety Stock
Imagine a product selling an average of ten units per day.
The supplier takes five days to replenish it.
And we've calculated that thirty units of safety stock provide the protection we need.
Our reorder point would be:
(10 × 5) + 30 = 80 units
When our inventory position reaches 80 units, we trigger replenishment.
Fifty units are intended to cover expected demand during the five-day lead time.
The remaining thirty provide protection if demand or supply doesn't behave as expected.
Of course, this is a simplified example. Actual replenishment systems may use different policies, and the inventory position must account for stock already on order and outstanding commitments.
But the principle is important.
Safety stock and the reorder point are not the same thing.
And changing demand can affect both.
What happens when demand changes?
Let's stay with our example.
When the reorder point was established, average demand was ten units per day.
But six months later, sales have increased.
The product is now averaging fifteen units per day.
The supplier's lead time is still five days.
If we leave the safety stock unchanged for a moment, the calculation becomes:
(15 × 5) + 30 = 105 units
That's a difference of twenty-five units.
The system previously waited until the inventory position reached 80 before triggering replenishment.
Under the updated demand assumption, it should trigger at 105.
And we haven't even recalculated the safety stock yet.
If demand has also become more variable, the buffer itself may need to change.
But it might not.
Perhaps demand has increased while becoming more consistent.
Perhaps the larger orders were exceptional.
Perhaps the supplier has become more reliable.
That's why we need to look at the evidence rather than simply increase stock whenever something goes wrong.
A stockout is a reason to investigate the inventory policy. It isn't, by itself, proof that the safety stock was wrong.
The calculation is only as good as the assumptions
This is where inventory management becomes more interesting than a formula in a spreadsheet.
Because the formula isn't really the difficult part.
The difficult part is understanding the behaviour behind the numbers.
What's happening to demand?
Are customers ordering more frequently?
Are individual orders becoming larger?
Are we seeing a genuine increase in underlying demand, or just a temporary spike?
How reliable are our suppliers?
Are the lead times recorded in the system still accurate?
How often do we review inventory and place orders?
And what level of service does the business actually expect?
That last question is particularly important.
Because two products with identical demand and lead times might reasonably have very different inventory policies.
One might be readily available from several suppliers.
Another might be difficult to source, with serious consequences if customers cannot obtain it.
Holding the same level of protection for both might not make much sense.
And the replenishment process matters too.
If we review stock and place orders once a month, we need to consider the demand that can occur before the next review, as well as during the replenishment lead time.
A calculation designed for continuous replenishment cannot simply be dropped into a periodic ordering process without considering that additional exposure.
There's a reason inventory planning involves more than setting a minimum quantity and hoping for the best.
When the system is doing exactly what we told it to do
I've always found this one of the more interesting aspects of supply chain systems.
We sometimes talk about an ERP system as though it makes decisions independently.
The system didn't order.
The system let the stock run out.
The system isn't working.
But often, the system is doing precisely what we've instructed it to do.
It's using the demand assumptions we've provided.
The lead times we've maintained.
The replenishment rules we've configured.
The safety-stock settings we've entered.
And if those assumptions are no longer appropriate, the resulting recommendations may no longer be appropriate either.
That doesn't mean the technology has failed.
It might mean the policy needs attention.
I think there's a tendency to treat inventory parameters as something we establish during implementation, revisit when there's a problem and otherwise leave alone.
But demand doesn't stand still.
Suppliers change.
Products move through different stages of their lifecycle.
Customer behaviour evolves.
And an inventory setting that made perfect sense twelve months ago might not be right today.
The system can keep using the same number long after the world around that number has changed.
Respond today. Improve tomorrow.
There's another part of this that I think is worth mentioning.
When something goes out of stock, customers don't particularly care how sophisticated our safety-stock methodology is.
They want to know when they'll receive their order.
And quite reasonably so.
That's why I think it's important to separate the immediate operational response from the longer-term improvement.
First, understand the shortage.
What orders are outstanding?
What stock is available?
What replenishment options do we have?
How quickly can we get product moving?
Then investigate the underlying cause.
Was it an exceptional demand event?
A supplier problem?
An incorrect lead time?
A replenishment rule?
Or an inventory policy that hasn't kept pace with changing demand?
Sometimes the answer will be to increase safety stock.
Sometimes it will be to change the reorder point.
Sometimes it will be to improve the forecast, shorten the lead time or change how frequently we replenish.
And sometimes we'll conclude that the existing policy was reasonable and the stockout was simply an event that fell outside the protection we'd chosen to provide.
That's an important outcome too.
Because a sensible inventory policy doesn't promise perfection.
It balances service, risk and cost.
The key is understanding what we're asking it to achieve.
A number isn't a policy
I think this is what stayed with me after that recent stockout.
A safety-stock figure might look like a simple number in an ERP system.
But behind it sits a whole collection of assumptions.
Demand.
Variability.
Lead time.
Service level.
Risk.
Working capital.
And ultimately, a decision about how much protection the business wants to provide.
Those assumptions deserve attention.
Not necessarily a complicated review every week.
But a sensible process for identifying when something has changed enough to warrant another look.
Because if demand has moved significantly and the replenishment settings haven't, perhaps we shouldn't be surprised when the results start to disappoint us.
And equally, if we're holding more inventory than the current demand and supply conditions justify, we should be prepared to challenge that too.
Good inventory management isn't about holding as much stock as possible.
It's about holding the right stock, for the right reasons.
And knowing when those reasons have changed.
Perhaps the real safety-stock illusion is believing that once we've calculated the number, the job is done.
Theme
Inventory Management
Signal
The danger isn't always getting the calculation wrong. It's assuming the calculation is still right.
Reflection
Safety stock is designed to protect against uncertainty, but its effectiveness depends on the assumptions behind it.
Demand changes. Lead times change. Supplier reliability changes. And the level of service a business needs may change too.
A safety-stock setting that was appropriate when it was introduced may no longer provide the intended protection.
That doesn't mean every stockout should result in more inventory.
It means we need to understand the relationship between demand, variability, replenishment lead times, service levels and the inventory policies we operate.
When a shortage occurs, the immediate priority is to respond to customers and restore supply.
But we should also take the opportunity to question whether the assumptions driving our replenishment decisions remain valid.
Because inventory planning isn't a calculation we complete once.
It's a policy we need to keep relevant.
Question
When did you last challenge the assumptions behind your safety-stock settings?
The Demand Signal
Lessons from the front line of supply chain.
