Well, supply chain is a competitive game. For a taste of that adversarial behaviors entail have a look at tv.lokad.com/journal/2021/3/31/adversarial-market-research-for-enterprise-software/ Cheers, Joannes
Hi, Joannes I have a quick question regarding the relationship between Available to Promise (ATP) and Safety Stock. When your Available to Promise (ATP) is 13 units and you are supposed to keep 15 units of Safety Stock. Does that mean you can still arrange those 13 units of ATP without dipping below your SS or you are supposed to keep 15 units of SS, which means you have to order more now to keep up with SS?
ATP >= LTD + SS where ATP = Available To Promise, LTD = Lead Time Demand, SS = Safety Stock When ATP = LTD + SS then you are supposed to reorder, you've reached the 'reorder point'. See www.lokad.com/calculate-safety-stocks-with-sales-forecasting However, safety stock is an outdated concept that should not be used anymore. Hope it helps. Cheers! Joannes
@@Lokad hey Joannes, I have to calculate a safety stock but I only have 3 months recent usage data and usage data from the same 3 months of the previous year. I'm given an example and asked to work others out . Example Feb 2020: 500 units, Jan 2020: 1000 units, Dec 2019: 750 units - then the previous year Feb 2019: 350 units, Jan 2019: 1010 units, Dec 2018: 500 units. Supplier gives a lead time of 90 days. The example gives a safety stock of 1000 units. Any idea how they came up with that safety stock figure?
@@neilgrigg7972 Overall, safety stock is an outdated supply chain approach tv.lokad.com/journal/2019/1/9/why-safety-stock-is-unsafe/ I do not recommend to ever use safety stocks outside a classroom ... As safety stocks depend on forecast, there is no *definitive* way to compute them. My take is that you have to figure out what was the intended way to build the monthly forecast baseline, and then apply www.lokad.com/calculate-safety-stocks-with-sales-forecasting Also, 90 days will probably be *assumed* to be 3 months. Best, Joannes
Demand is not normally distributed (aka not a Gaussian), thus, the first step to start improving upon the safety stock consist on using an alternative approach. The safety stock *formula* cannot be fixed because the underlying approach is broken by design. This very insight cost me the first 4 years of Lokad or so. Historically, we found that quantile forecasts represent very significant improvement over safety stocks. Later, we realized that probabilistic forecasts were a big improvement over quantile forecasts. Hope it helps, Joannes Vermorel
![Kodak Black - Versatile 4 [Official Video]](http://i.ytimg.com/vi/VcP_1Ua__vY/mqdefault.jpg)
Rob Stark figuring out the best way to get supplies from the North
Well, supply chain is a competitive game. For a taste of that adversarial behaviors entail have a look at tv.lokad.com/journal/2021/3/31/adversarial-market-research-for-enterprise-software/ Cheers, Joannes
Hi, Joannes I have a quick question regarding the relationship between Available to Promise (ATP) and Safety Stock. When your Available to Promise (ATP) is 13 units and you are supposed to keep 15 units of Safety Stock. Does that mean you can still arrange those 13 units of ATP without dipping below your SS or you are supposed to keep 15 units of SS, which means you have to order more now to keep up with SS?
ATP >= LTD + SS where ATP = Available To Promise, LTD = Lead Time Demand, SS = Safety Stock
When ATP = LTD + SS then you are supposed to reorder, you've reached the 'reorder point'.
See www.lokad.com/calculate-safety-stocks-with-sales-forecasting
However, safety stock is an outdated concept that should not be used anymore. Hope it helps. Cheers! Joannes
Hello lokafd, is there anyway you could help me with something that I can't really explain on here?
Ask away :-) I can promise an answer, but I will certainly try. Cheers. Joannes
@@Lokad hey Joannes, I have to calculate a safety stock but I only have 3 months recent usage data and usage data from the same 3 months of the previous year. I'm given an example and asked to work others out . Example Feb 2020: 500 units, Jan 2020: 1000 units, Dec 2019: 750 units - then the previous year Feb 2019: 350 units, Jan 2019: 1010 units, Dec 2018: 500 units. Supplier gives a lead time of 90 days. The example gives a safety stock of 1000 units. Any idea how they came up with that safety stock figure?
@@neilgrigg7972 Overall, safety stock is an outdated supply chain approach tv.lokad.com/journal/2019/1/9/why-safety-stock-is-unsafe/ I do not recommend to ever use safety stocks outside a classroom ... As safety stocks depend on forecast, there is no *definitive* way to compute them. My take is that you have to figure out what was the intended way to build the monthly forecast baseline, and then apply
www.lokad.com/calculate-safety-stocks-with-sales-forecasting Also, 90 days will probably be *assumed* to be 3 months. Best, Joannes
What is your recommendation to improve the safety stock formula based on gausian distribution?
Demand is not normally distributed (aka not a Gaussian), thus, the first step to start improving upon the safety stock consist on using an alternative approach. The safety stock *formula* cannot be fixed because the underlying approach is broken by design. This very insight cost me the first 4 years of Lokad or so. Historically, we found that quantile forecasts represent very significant improvement over safety stocks. Later, we realized that probabilistic forecasts were a big improvement over quantile forecasts. Hope it helps, Joannes Vermorel