Grocery delivery e-commerce: the technical side
Groceries are the category where technical mistakes cost the most: spoiled goods and a late order can't be undone after the fact. Here is what's specific to it.
Real-time stock is not optional
Unlike clothing or electronics, where a stock mismatch just delays an order, in groceries it means the order physically cannot be fulfilled. Sync needs to be near real-time, not hourly.
The sensible setup is tight integration with the inventory system of the specific location fulfilling the order, not a combined stock figure across all locations.
Delivery zones and time slots as the core order logic
Grocery delivery hinges on specific time slots and a radius from the fulfillment point. The site should show available slots for the customer's address before checkout, not after.
Outside the delivery zone, the cart should not allow checkout at all — otherwise a manager ends up manually cancelling orders, which kills customer trust.
Weight-based and unit items
Some products sell by weight rather than by unit, and the cart price should be recalculated after actual weighing during fulfillment — the site needs to support 'estimated price / final price after fulfillment' logic.
Substituting an out-of-stock item requires customer consent: the best practice is offering 'substitute with similar', 'remove from order' or 'confirm with operator' directly in the interface.
Speed as the key conversion factor
Site speed is doubly critical in this category: grocery orders are often placed in a hurry before cooking, and slow catalog loading directly costs orders.
Reorder ('order like last time') is one of the most effective features here — grocery shoppers tend to repeat the same list week after week.