Wouldn’t it be nice if calculating safety stock wasn’t such a crap shoot? The biggest conundrum in demand planning is Service Level Management and how to calculate safety stock so that the company can meet its service level goals — without sabotaging margins, that is. That’s the topic of today’s post from our ‘7 Essential Steps’ series.
In case you were sleeping, we’ve already gone through Demand Forecasting, Lead Time Forecasting and Order Cycle Optimization. Here in Step #4, we’ll share some proven Service Level Management strategies that will help bring greater accuracy to your demand forecasts, so you can hit the highest service level goals possible for your company’s investment.
Getting Real with Safety Stock Calculations
Buffer stock, extra stock, C.Y.A. stock… even overstock are names fondly given to Safety Stock. And to a certain extent, each name describes safety stock fairly accurately. Each name conveys the fact that extra on-hand inventory is involved. However, none of the terms adequately describe what the extra inventory is for, with the exception of C.Y.A. Stock, of course. But even C.Y.A. stock misses the whole point of investing in extra inventory. SAFELY.
No Demand Forecast is Perfect
Let’s just clear the air right up front. To think you can have zero safety stock is Fantasy Land.
When we speak of safety stock, we often talk in terms of “inventory needed to protect the service level goal,” or “a level of extra stock that is maintained to mitigate risk of stockouts.” This description isn’t wrong, but it doesn’t convey the real purpose of safety stock. Until some genius (we’ve got a couple of them here… and they say they are close!) develops what we like to call the “Crystal Ball” demand forecasting algorithm, at least some amount of safety stock is going to be needed to protect against the error that exists with demand forecasting.
I hate to be the bearer of bad news, but while some forecasting algorithms are better than others in certain circumstances, none are 100% correct every single time. It is those forecasting errors which we need to account for.
Supply Chain Planning for Inventory Control
One of the best ways to account for forecast error is obviously to have a demand planning process by which the forecast is maintained regularly, applying seasonal profiles when needed, and making proper adjustments when called for. Supply chain planning (SCP) algorithms have the power to calculate a safe amount of inventory to protect against the inherent forecast error you face every day.
Because inventory control is always an initiative, and options exist in SCA to ‘control’ safety stock (manual safety stock, safety stock min., safety stock max.), many demand planners will ask: “Should I change the safety stock quantities in SCP?”
Inventory planners should understand that the options are there because every environment is different, and there is always an item or two that defies typical best-practice principles. But in general the answer is ‘No’. Safety stock is one of the heaviest, most complicated calculations in the system. And it should be recalculated every day.
5 Safety Stock Calculations that Count
If a demand planner feels that too much is being carried in terms of safety stock, demand forecasting technology can be used to optimize the following 5 components of calculating safety stock:
- Demand Forecast (demand velocity) – is the forecast correct? Too high a forecast will drive up safety stock!
- Standard Deviation (a measurement of forecast error). This is one of the main drivers of safety stock. The greater the deviation, the greater the need for safety stock. You may not feel like you can control this component much. However, what you can do is use technology to find seasonal tendencies and unfiltered promotional demand. Here are some key tips to follow here:
- Applying an appropriate seasonal profile will typically lower you standard deviation and your safety stock needs.
- Look at the spikes in the demand you have. Can you attribute any of these spikes to promotional demand (which hasn’t already been filtered)?
- Keep in mind that, typically, the slower the demand, the greater the standard deviation. It’s normal. With that in mind, you could have a high standard deviation with an item you feel you can lower. Keep in mind that even with the high standard deviation for the slow moving items, forecast is still being used. So while you may have a ‘high number of days safety stock,’ more than likely, you won’t have a high quantity of safety stock.
- Lead Time and Lead Time Deviation. The greater your lead time, the more opportunity you have to miss sales (provided that you aren’t overstocked!). Supply chain planning solutions understand that and use Lead Time as a component of safety stock. Check your lead times and make sure they are accurate. Lead time variability works the same as the standard deviation component. If you have vendors that are consistent in their deliveries, on time and complete, then you can afford to have lower lead time variability. If not, you may need even greater lead time variability.
- Order Cycle – Yes, even your Order Cycle is used. This component actually reduces safety stock as the order cycle increases. Think of it in this way… If you have an Order Cycle of 3 days, you are only going to carry 3 days of Order Cycle inventory. A demand spike could put the ability to ship in jeopardy. On the other hand, if your Order Cycle is 180 days, you are carrying more Order Cycle inventory and a spike is less likely to affect your service.
- Service Goal – This is arguably the main driver of safety stock. And, quite simply, if you feel you are carrying too much safety stock, you need to ask: “Is my service level too high?” If it isn’t, you probably really need the safety stock!
There are many Service Level Management strategies that can be employed, and demand planning tools that can be taken advantage of when looking to ‘control’ inventory levels. These tools should be in your arsenal, 100%.
Want to learn more about advanced supply chain planning tools for accurately calculating safety stock? Reach out now and we will be happy to develop a best practices strategy with you!
Be sure to subscribe to PlanningPosts to follow this series.
Back to Part 1 – Demand Forecasting
Back to Part 2 – Lead Time Forecasting