Retail in the Age of Disruption (Part Three)

This is the third entry in the CityStream Solutions blog series discussing Four ways that Big Data, Data Science, and Predictive Analytics must evolve and innovate. Navigate all Age of Disruption series entries: (1) | (2) | (3) | (4) | (5)

From the “efficient” warehouse model to the “agile” warehouse model

The most efficient distribution network is the least flexible. Retailers who over-adapt to a temporary “normal” will be exposed when the next major disruption arrives. For example, legacy department store chains that neglected to develop their online browsing and purchase capabilities found themselves suddenly cut off from their customers, while brands with fully developed e-commerce sites were able to better capture revenue during the shutdown.

Amazon’s warehouses are gigantic, always filled, and ruthlessly efficient. The company set the expectation of 2-day and sometimes 1-day free shipping, which seemed impossible to compete with. But the pandemic has exposed the brittle nature of their brutal efficiency, as Amazon quickly abandoned 2-day and 1-day shipping and started prioritizing “essential” items over “non-essential.” Discouraging people from demanding free 2-day shipping on video games in favor of acquiring and delivering more toilet paper and vital food products was the right thing to do. But it does show retailers that there is a path forward, even up against a goliath like Amazon. It only took two weeks to show that the “get anything you could want, fast” e-tailer was overfit to a consumer world that no longer exists.

Let’s take a hypothetical disruption like a tainted water supply, where people can’t trust the tap water (maybe it’s a water-borne pathogen, an act of bioterrorism, or lead in the pipes). Suddenly, there’s demand for bottled water all across the country, all at once. In this example, bottled water is Product X. Assuming the manufacturing sector can keep producing a lot of water bottles very quickly, the question for distributors and retailers is how to get to get it to the customers who desperately need it, especially as the current supply disappears from supermarket shelves. There must be available space in the warehouses ready to handle the influx of Product X (this time, water bottles). But it’s considered inefficient to keep sections of warehouses empty just in case something unexpected happens. How to balance these two opposing forces?

Analysts must create models in advance of disruptions, for how various products can quickly take over a lot of warehouse space, and how other products can be moved, discounted, and sold to make room for them. We must leverage the shelf-by-shelf and product-by-product data being generated as algorithms track the movement of products and changes in demand. Our new generation of predictive analytics must include multi-location models. Where we need shelf space, we must be agile in moving its occupants elsewhere, including the cost of trucking it to holding locations.

By modeling and planning for potential disruptions before they occur, a company can seize the speed advantage. When a disruption arrives and it becomes clear which product(s) is going to be Product X for this round, a firm can immediately instigate its multi-location redistribution plan, executing a pivot while competitors are just realizing they have too much of some things, not enough of Product X, and no idea how to make room for it. Data-driven solutions will become even more critical to company success.

Continue on to part four here.

Questions? Comments? Contact us.