The first wave of eGrocers in the late 1990s like Webvan, failed to garner significant market share despite heavy investment in automation of their warehouse operations. The typical investment in such automation ranged from 4-8 x that of ordinary warehouse (owned / leased) and the key was not necessarily throughput but efficient utilization of assets, while adopting a cost-efficient delivery and service level (traditional tradeoff of efficiency vs customer service).
One of the methods of improving asset utilization is through batching where orders are consolidated into “batches to be delivered at particular time windows” – broadly, early morning (before work), late morning / early afternoon (work from home or with retired people at home to take delivery) and late evening(after work). Optimal number of batches can be simulated through 3D discrete time event simulation package based on input parameters (customer numbers, preferences, product range, etc.) and adoption of Lean practices. (Source: Simulation of Outbound Operations in an eGrocer Warehouse, Gustavo Costa, MIT, Jun 2018)
Increasing batching also significantly reduces WIP inventory, improves area and asset utilization, while adversely impacting picking and delivery costs (viz. higher cost due to sub-optimal picking paths and delivery routes).
E-Grocers make decisions on their operational model based on focus (Online, Omnichannel), storage (dedicated warehouse/VMI or outsourced) and delivery mode (door delivery -attended / unattended, order pick up – store / kerb). Players like Instacart completely forgo physical locations and fulfil through brick and mortar stores operated by partners.
While dedicated warehouses increase fixed costs, they create a risk pooling effect and thereby reduce demand variability (reduction of the square root of # customers for same skew or by square root of skews for a given product), FIFO at scale thereby minimizing spoilage as well as reducing any duplication in stowing and picking costs.
Our experience with Indian eGrocers is the need to explore batching as a way to reduce total supply chain cost (reduce inventory holding and spoilage offset partly by increased freight cost), leverage low cost automation (labelling and scanning for faster, error free data entry to the lowest SKU), belt conveyor systems for material flow, rack stacking flexibility through periodic data entry into WMS/ERP and process improvements for optimal working in light of varying order volumes and ever-increasing brand and line product mix.
Focused solutioning providers like Visilogix (www.visilogix.com) can help.