In today’s retail industry, change is the new normal. There are three key factors driving this change including: dramatically changing customer preferences, a major shift towards omnichannel purchases, and an increasing need for retailers to respond quicker to shifting preferences and buying behaviors with more efficient sourcing, transportation and fulfillment.
This and other insights about how retailers can bolster their approach to omnichannel challenges were shared in “Fulfillment Planning in an Omnichannel World” featuring:
- Logan Vadivelu – Global Lead Industry Solutions – Retail at Google
- Vikram Murthi – VP Industry Strategy at o9 Solutions
- Peter Taylor – Presales Director at o9 Solutions
Here are four takeaways for retailers to keep in mind as they develop an omnichannel fulfillment planning strategy.
Increasing demand volatility continues to drive omnichannel challenges
During the pandemic, lockdowns and public safety measures resulted in a significant drop in in-store retail purchases and drove new customer demand patterns and a surge in omnichannel purchases.
Many retailers experienced increased demand volatility and were challenged by the increased order volatility from eCommerce channels, increased promotional activity, shorter product life cycles coupled with regional variances, and rapidly changing assortments.
Demand variability and forecast accuracy is a long-standing challenge. As the pandemic set in, many retailers realized that historical data is not a predictor of future demand and that other external factors, such as local events, play a significant role as well. Aggregating past history and leveraging diverse external data is a key component to improving the accuracy of demand sensing capabilities.
“Combining those signals becomes critical in terms of improving accuracy from a demand sensing point of view,” says Vadivelu. “There are data points that talk about how 25 percent of shoppers are going to continue to do online shopping—that’s a major shift in behavior.”
Retailers are relying on more external data points to address demand volatility
As more consumers embrace omnichannel purchasing, retailers are keeping up with this shift in behavior by incorporating more local data into their demand sensing and running forecasts more often to ensure consumer needs are met. Both machine learning and deep learning models have become a critical component in being able to run forecasts more frequently with greater accuracy in order to stay on top of shifting consumer demands.
How retailers are fulfilling product availability efficiently
The pandemic also put pressure on retailers to maintain product availability and deal with increased expectations for online and on-shelf available in the face of severe shortages or overstocks due to demand peaks and slumps, potential supplier shortages, and frequent port and transport lane shutdowns.
As a result of these constraints, retailers are changing how they approach omnichannel fulfillment. The process is becoming less siloed between in-store and online fulfillment and transitioning towards an all-encompassing approach where, depending on availability, fulfillment can be shipped from a distribution center or store directly to a customer. “What retailers realized is that it’s better to have a holistic inventory policy and to leverage stores as shipping locations,” Murthi says. “A great example of this is Target. They realized they can’t open a centralized e-commerce fulfillment center fast enough, but had thousands of stores, so why not leverage that. It’s good for the consumer but also great from an efficiency standpoint for retailers.”
The role of machine learning in omnichannel fulfillment
Machine learning capabilities have played a significant role in retailers’ ability to quickly pivot and effectively handle omnichannel fulfillment during the pandemic and beyond. Both ML and deep learning capabilities helped retailers to:
- Sense the demand and store sell through and gain more real-time visibility into the omnichannel space
- Quickly bring in market drivers (weather, mobility indexes, Covid infection rates, etc.) that were cleansed and harmonized to provide visibility into external factors that were directly affecting consumer behavior
- Create forecasts at a granular level to drive effective inventory strategies to best handle demand volatility
For retailers, having the ability to not only access external data points, but to be able to apply this knowledge to create more accurate forecasting models allowed them to respond to shifting demand and potential disruptions within their supply chain as efficiently as possible.
“That’s where having a real-time aspect is critical because you are able to apply these things that influence the forecast results. It’s truly reflecting the demand volatility in your forecast prediction,” says Vadivelu. “ML and other approaches are important, and the ability to bring in data at the right time is critical to influence the forecast. Otherwise, you mist the window to influence that signal.”
o9’s Control Tower can help retailers incorporate accurate real-time data into their planning
The webinar concluded with a demonstration of the o9 platform’s retail control tower that included:
- Demand sensing with market knowledge
- Omnichannel inventory placement
- Omnichannel product availability
Many retailers are trying to determine where to place their inventory and often lack sufficient, real-time data to do so efficiently. The o9 platform incorporates external data, powered by Google, to provide insights into demand volatility, allowing retailers to respond accordingly.
Peter Taylor, o9’s Presales Director, showcased an example of how retailers in France were affected by Covid’s impact on consumer mobility. His graph showed how Covid spikes affected consumer’s buying patterns and how retailers can use a control tower to translate demand and supply signals into an accurate supply plan. The processes can be automated to alert retailers to where bottlenecks exist and make recommendations on how to best manage inventory.
“We can solve constraint issues and check financial implications to see exactly how in-stock percentages, revenue and margins are affected by a scenario,” Taylor says. “You can really review the KPIs in terms of revenue, in terms of margin, logistics costs and expedites to say, ‘I’ve quickly found a way to resolve the issue,’ but also within the context of maintaining costs within the supply chain.”
Overall, while retailers continue to be challenged by increased demand volatility and product availability, external data points can help them to build out a robust forecasting strategy that allows retailers to proactively respond to supply chain disruptions.