I'm starting a discovery process on the requirements -- operational, systemic, planning/forecasting, etc. -- to opening a forward-deployed inventory location outside of a major city, with the primary purposes are rapid owned retail store replenishment (to alleviate limited back-of-house availability) and local e-commerce order fulfillment. Can you share your experience in the requirements to bring this type of operation to fruition?

Range management i.e. the SKU range will be different from a larger major city, so you may have usual runners/repeaters but not the complete category (I think that point made in other comments).  Also, that you will need to consider localised demand that is diluted when you look at national / larger city demand, so again will contribute some SKUs to the range that you might no ordinarily consider.

Forecasting and safety stocks as presume again smaller location so physical limitations apply and to the same point, replenishment frequency is important

Ability to cross dock in from NDC into stock location for picks to serve ecommerce orders that are maybe out-with stocked range

Reverse logistics and processes associated with returns management as presume limited space.

PLANNING, PLANNING, PLANNING.  The issue with ecommerce demand profiles is the SKUs that make up the bulk of the demand aren’t usually what you’d use to stock a traditional store.  For example, in apparel, your top movers may be socks, underwear, and accessories where a traditional planning model would try to put complete categories.

If the inventory is solely stored in the back and is used for ecommerce, I’d begin by getting the top selling 10% of SKUs in that market.  This is also tied to your orders’ sourcing logic.

Your goal should be to try and create a node that has a wide breadth of assortment in limited quantities so you maximize the chance of completing that order from that node.