The Innovator's Solution


Supply Chain Innovation: Adapting Demand Forecasting to Multichannel

Posted by Jeff Bodenstab on Jun 2, 2015 2:00:00 PM

Multi-channel_shopping1Multi-channel_shopping3Last week we described how Gartner analysts Matthew Spooner and Tom Enright expected supply chain planning to evolve between now and 2025. This week we focus in on how they expect demand forecasting to adapt to multichannel.

According to Enright, multichannel needs a new way of forecasting demand – rather than just answering the question “How many will we sell?” – also addressing the question “How will the consumer buy?”  The problem is that most multichannel demand forecasting today is still one-dimensional, focused on product demand per week, for weeks or months into the future. However, the processes and resources to meet this demand vary according to how the consumer actually makes the purchase.

Besides buying on-line or purchasing in store, there are many other combinations of ways consumers can shop and have their orders fulfilled, including:

  • Click-and-collect in-store (buys online and collects within the store)

  • Click-and-collect remotely to the store (buys online, and then collects from a locker location or drive-through collection outside the store)

  • Reserve online and collect in store (order online, pay and collect within the store)

  • Buy online and ship from store to their home or another address 

Forecasting each channel enables an approach where instead of always reacting to demand with expensive cross-channel fulfillment, the fulfillment channel is planned proactively before getting the customer order. This way, all components of the chain have “advanced visibility into the anticipated fulfillment volumes for each method and be able to resource and plan accordingly to achieve higher on-time fulfillment rates,” Enright says.

To do this, modern supply chain planning systems must forecast this multichannel demand at the transactional level, Enright says. They must calculate the demand forecast by processing data about how the customer completed the purchase. This includes, Gartner says, product information, down to color and size; transaction date; store location credited with the transaction; where the transaction was initiated (in-store or online); how the order was fulfilled—picked in-store, picked in distribution center, picked by supplier; and how the consumer received the product—in-store collection, shipped from distribution center, shipped from supplier.

Retailers and suppliers need to collaborate closely in multichannel fulfillment. Customers don’t care if items are delivered by different suppliers, as long as they can order in a straightforward manner and get delivery in a hassle-free way. “They don’t want 20 different deliveries to their home; they want to receive a single delivery,” says Spooner.

Returns forecasting is also crucial. Returns average 25% in clothing and up to 45% in categories like dresses, men’s suits, and shoes—items then sold at substantial discount and loss of profitability. Monitoring and forecasting returns by channel, product, and customer segment—using predictive analytics to understand the causes—is imperative for profitable supply chain planning.

“It’s about being able to respond profitably; this is where planning comes in,” Spooner says. “It’s no good having all of this capability if we can’t deliver it profitably.”

For a case study about how one apparel company addressed  the challenges of multichannel demand, click on the case study below:

Pompea Case Study


Topics: Forecasting Demand and Analytics, Supply Chain Planning, Retail and Multichannel

Share this article or your comments below: