In D2C Commerce, Every Second Counts

Amazon robots move inventory for pick, pack, and shipment.

Amazon robots move inventory for pick, pack, and shipment.

In September, we talked about innovative brands using predictive algorithms to transform customer experience. This month we evaluate the importance of real-time data processing toward the same end goal; to build customer equity (the number one driver of brand valuation) by creating transformative customer experience.

Just yesterday, The Wall Street Journal broke the story on Amazon's plans to expand the cashier-less Go concept to Whole Foods and other large-format retail stores. Retailing headlines focus on Amazon’s IoT, third-party seller competition, product assortment, distribution center footprint and fulfillment automation as competitive advantages, but the core capability driving the commerce juggernaut and its customer experience is real-time data processing.

How can real-time data transform D2C customer experience to grow customer equity? Using real-time data to inform predictive algorithms means more accurate modeling and forecasting. Here are a few use cases, from basic to complex:

  • Flag orders for warehouse or customer service intervention before an exception occurs

  • Benchmark intraday metrics to revise operational and financial forecasts to plan in real time. At 9:30am, are we on target to achieve our daily goal?

  • Monitor secondary signals of operational exceptions such as call center volumes and customer cancellation rates to identify root cause exceptions. Use the early warning signs as training data to predict future exceptions

  • Flag open orders that will miss an upcoming scheduled carrier pickup from the warehouse. Trigger a capacity change, or upgrade to expedited shipping and notify the customer of the change, reiterating on time delivery

  • Initiate an inventory transfer from one channel (ecommerce) to another (Amazon), or one geography (west coast) to another (east coast) based on early demand indicators

  • Flag a fulfillment capacity or available-to-sell (ATS) inventory constraint based on social media activity or traffic to a specific product detail page (PDP)

  • Calculate unit quantities for supplier replenishment P.O. based on predicted sales velocity and available inventory

The full potential of real-time data lies in powering predictive analytics and the decisions and actions they support. In each of the these cases, it is much better to predict in advance the exception, variance, missed shipment, capacity constraint, or quantity based on a high confidence forecast, updated live with the latest operational and customer developments.

Eric Best