Market-based demand forecasting to help you predict orders and resources

Retailers are facing increasing pressure to offer affordable, reliable and frictionless omni-channel fulfillment experiences. With the acceleration of omni-channel shopping, businesses must be able to predict labor needs, improve on-time delivery and order fill rates, and create engaging and convenient store pick-up experiences.

One large grocery chain in North America faced challenges due to increases in omni-channel shopping. Poor customer experiences were being created due to out-of-stock items, crowded aisles, and insufficient resources to complete in-store fulfillment. To address these issues, the company implemented a first-of-its-kind solution for Fulfillment Order Forecasting; now the company can create order forecasts by hour and by store, and with the help of machine learning (ML), can predict online demand. So far, the grocery giant is seeing a 15% reduction in overtime, a 25% reduction in daily forecast errors, and a 95% increase in customer satisfaction.

This customer example is just the beginning of a whole new way of retail and grocery planning solutions that we call Fulfillment Forecasting. To explore these topics more, I recently hosted a Blue Yonder Live session with Erin Halka, Senior Director Solution Strategy, and Badri Krishnamachari, Corporate Vice President, Retail Solutions. Read on below for a Part 1 summary of the conversation and then go hear it straight from them on the Blue Yonder Live.

Market-based Demand Forecasting

According to Badri Krishnamachari, “The key to successful fulfillment forecasting and successful omni-channel operations is really to get a good idea of what a demand would look like through unconstrained demand.”

Badri also shares that the key to success lies in truly understanding demand, the fulfillment options available, and the inventory and resources needed to meet that demand. Not just looking at historical demand, trends and data, but also considering future fulfillment strategies that could involve meeting demand from different locations or opening up new stores or micro-fulfillment centers.

Understanding unconstrained demand involves considering assets such as labor needs and resources (ex., totes, refrigerators, trucks, etc.) that are dependent on order volume.

By having better predictions of labor needs to satisfy demand, companies can better schedule staff members and understand the tasks they will be doing while working. This leads to better job satisfaction and cost savings as companies deal with rising prices due to inflation.

The Importance of Resource Placement

With the volatility of how the customer choses to get their online order – pick-up/curbside or delivery –  it is crucial for businesses to have better predictions to identify the customer orders and the resources need to avoid missing sales or having excess labor or even worse, overtime, that leads to margin erosion.

According to Erin Halka, “If you have better predictions in what those labor needs are going to be to help satisfy that demand, it allows you to be able to better schedule those staff members. It allows you to be able to understand the tasks and the things that they’re going to be doing while they’re working that day, and it leads to better job satisfaction. Ultimately, this also helps with cost savings as the top trend companies are dealing with prices going up with inflation.

One challenge is the regionalization effect that is happening. Companies need to think about how they can better service the market and make sure they have the ideal time slots available to service customers, be it a store or micro-fulfillment center (MFC).

Another challenge is the change in shopping behavior. Companies need to shift their thinking away from location-specific demand to market and customer-centric demand and assign that demand to the locations and channels that can fulfill it. ML can help with this by picking up promotional and causal influences that are local in nature.

Traditional approaches to planning demand, supply or assets tended to be fixated with location. However, this can be misleading as fulfilling locations can be dynamic in the case of e-commerce. Companies need to move away from traditional methods and embrace ML to better manage their resources.

Optimizing Resource Placement with Blue Yonder

Predicting future customer orders to guide placement decisioning is crucial, as is having the flexibility for decisioning that depends on the business needs. This means companies need to prioritize intelligence in guiding where and which locations to place resources, based on current market demand.

With Blue Yonder’s Fulfillment Forecasting microservices, companies can optimize the total cost of omni-fulfillment by considering various cost factors, including labor, logistics, capacity, markdowns, and stockouts. By creating a market-level, e-commerce forecast and dynamically apportioning the forecast to ideal nodes while respecting capacity constraints and fulfillment rules, companies can enable better order and resource predictions placement that plugs into existing Workforce Management (WFM) systems without any changes needed.

With optimized fulfillment and labor forecasting, companies can predict labor and hard asset needs to process e-commerce orders. By defining the resources needed to guarantee perfect availability and forecasting their consumption over time, companies can align assets to omni-channel demand at ideal fulfillment locations for their consumer’s convenience. This saves time, capital and carbon while protecting walk-in sales.

Learn more by listening to the conversation on the Blue Yonder Live. The Part 2 blog post will be coming soon.

Learn more about Fulfillment Order Forecasting and Fulfillment Item Forecasting solutions.