The incumbent approach to store fulfilment was to allocate orders to stores as quickly as they come in so that they get to stores as quickly as possible, this is known as push-based fulfilment. This approach does not work at scale because orders can come in quicker than stores can pick them, and the picking process starts to fall behind. As a result, the dispatch point becomes overloaded, and the team starts to struggle. While they are trying to catch up, time passes, and the store stock is being sold off the floor. Furthermore, while one store is being allocated more orders than they can handle, other stores that could have helped are ignored- a massively inefficient use of resources. Worst of all, stock is being sold out after allocation results in short picks, order voids and customer dis-satisfaction.

Viare’s pull based fulfilment model allocates orders that need to be fulfilled to a “global pool” and the actions of the store would drive the decision process and stock to be allocated for picking in real time. Effectively this means that a store would be allocated an order to be picked at the time that they were ready to pick that order.

Ultimately, pull based fulfilment mitigated the risk of picking within a live retail environment- especially during peak trade. As a result, voids reduced from 10% to 0.5% (including orders that customers cancel themselves) almost immediately.