Data
Set reorder policies
Safety stock and reorder points per SKU, from actual demand variability.
Fill it in
By SKU over time.
The fill rate you want.
One policy across a mixed catalogue is wrong everywhere.
Your prompt
Set the policies. Service level: [your service level] Compute safety stock from the VARIABILITY of demand and of lead time together. Using average demand times a fixed number of weeks is the common approach and it over-stocks the steady items while starving the erratic ones, which is exactly backwards. Segment first. A single policy across a mixed catalogue is wrong everywhere: fast steady movers, seasonal items, and slow erratic ones need different models, and intermittent demand needs a different method entirely. Show the cost of the service level, in stock value. Going from 95% to 99% is often a doubling, and nobody makes that choice consciously. Flag the SKUs where history is too short to conclude anything.
Use Set reorder policiesOpens with everything above already filled in.
Why this works
Average demand times a fixed number of weeks over-stocks the steady items and starves the erratic ones, which is exactly backwards. This computes from demand and lead-time variability together, segments first, and prices the service level in stock value.