Retail Trends

Retail competition has largely been driven by scale efficiencies in purchasing, supply chain, and retail. It began as individual proprietor stores and restaurants that eventually gave way to “chain stores.” Smaller chains then gave way to big box retail chains as they continued to aggregate scale and leverage more efficient larger store formats. Retail chains themselves are now giving way to digital marketplaces, with leaders wielding unprecedented scale and eliminating retail store overhead altogether.

The devastating pandemic accelerated eCommerce growth, which is expected to be 21.8% of retail by 2024.1 It forced retailers to rapidly implement safe distancing capabilities—buy online pick up in store (BOPIS), curbside pickup, and store-to-home delivery—and redesign their Distribution Centers (DCs) to accommodate the higher volume of eCommerce orders. Most retailers missed last season’s peak order volumes; many are struggling to bring capacity online for this season. Sourcing delays and supply chain capacity shortfalls are expected to continue through this holiday season. All are struggling to find innovative ways to stop margin erosion from handling more costly eCommerce orders.

Today, eCommerce delivery service competition is fierce. Free next-day, same-day, and two-hour home delivery windows are becoming the new normal, virtually eliminating consumer trips to local stores. ECommerce leaders now have enough scale to build large and efficient distribution centers in local markets,2 replacing as many local market stores as consumer buying patterns allow. These DCs—some measuring an impressive 300,000 square feet in size—are being built with advanced material handling automation to reduce both costs and dependence on labor. Likewise, these leaders have dramatically lowered last-mile delivery costs. For example, Amazon already averages 20 stops per hour (or 200 stops per 10-hour shift) and will improve this with the new larger capacity UPS-style vans being added to their fleet.3 This unprecedented scale and efficiency in the local market coupled with free next- or same-day delivery is a fierce competitive strategy.

Competition is also escalating for the re- tail storefront of the future, as consumers consider their shopping preferences (online only vs. an online/in-store mix). Amazon is already rolling out its Go and Fresh stores, which have eliminated cashiers, the hassle of waiting in line, self-scanning, and payment—all dramatically reducing the consumer’s time in store. Similar initiatives are underway with leading retailers, many eager to leverage advanced technology to reduce store labor and streamline the consumer experience. Competition is fierce on all fronts.

Continuous Innovation

This overwhelming scale, pace, and complexity of change, coupled with consumers expecting free next- and same-day delivery, is requiring retailers to completely overhaul and innovate the entire way they used to get products to consumers. The days are over of developing a five-year supply chain network strategy that focuses on where to place our 2 eCommerce DC's and our 5 to 12-store replenishment DC's. It has become a question of how we can build micro-fulfillment centers in every local market to deliver next- and same-day delivery without eroding margin (i.e., going out of business). In other words, today’s continually mounting competition requires a radical and ongoing network innovation response. The focus of this paper is on innovative ways to model new local network scenarios.

The Four Locals

Thus far, leading retailers are responding by innovating local market networks in four key areas—called the four locals. These are: adding new highly automated fulfillment centers and delivery fleets in local markets, redesigning inventory strategies to optimize local fulfillment, and aligning with partners to aggregate scale to lower local fulfillment and last-mile delivery costs. The following are some examples of what leading retailers are doing:

Amazon already has 110 local market fulfillment centers and dozens more on the way to meet its goal of next- or same-day delivery.4 Many of the largest retailers with deep pockets and wide margins appear to be matching Amazon’s service strategy by adding networks from 50 to 100 or more local fulfillment DC's of their own. Retailers that benefit by specializing DC's according to material handling method (e.g., conveyable or non-conveyable) plan to add even more. When we factor in all the other DC's needed to support fulfillment— import, inbound sort, outbound sort, delivery hubs, and returns—the sheer number of new DC's being built today is staggering. It’s no wonder space and labor costs are on the rise and in short supply, driving explosive investment in material handling automation.

Sources: www.Mwpvl.com

The Four Locals

Similarly, Home Depot is building 40 new flatbed fulfillment DC's for large orders, and another 50 DC's are on the books for general merchandise orders.5 Big box retailers such as Sam’s Club have already converted key store locations into fulfillment centers with advanced automation and next- or same-day delivery capability. Other retailers are testing new hybrid concepts by adding micro-fulfillment centers in store backrooms or on mezzanines, increasing stock availability for in-store shoppers. Walmart and other retailers are already signing contracts to delivering products for other retailers to aggregate scale, lower costs, and capture last-mile revenue.

Likewise, less massive retailers with shallower pockets are experimenting with a similar strategy, albeit on a smaller scale. They are turning select stores and new unoccupied low-cost mall locations into highly automated micro-fulfillment centers with delivery capacity. These retailers are more carefully crafting their strategies as they don’t have the scale or resources to build 100 or more local market DC's. Accordingly, they are carefully selecting the areas for which they can and should offer expedited delivery service. In our opinion, this is the true battleground at which Amazon is taking aim: retailers who don’t have the scale to compete in a free next- or same-day world. Amazon’s gamble is that what they lose in free shipping, they will more than makeup for with new business. These are just a few examples of the mounting retailer network innovation response. State-of-the-art ideas are on the drawing board—and need to be for survival in the emerging new era of retail.

Source: homedepot.com

How to Continually Innovate

The question then becomes: What’s the best way to go about modeling these new local network scenarios? Traditional network modeling software is great at modeling national and regional networks. They accurately estimate the differences in distribution center labor and space cost by region of the country. They also calculate the transportation cost differences for each DC scenario, using carrier point-to-point rate tables for each mode: truckload, less than truckload, and parcel shipments.

Why can’t most traditional network modeling tools be used to solve the local network design problem? As discussed, the new era of free next- or same-day delivery and the concept of operating a hundred or more local DCs was unheard of just a few years ago. Therefore, most tools used today were not designed to meet this requirement. Further, much of the functionality they do have is not applicable to local market design. For example, the ability to estimate DC costs between regions is not needed, as labor and space costs don’t change in the local market. Parcel carrier rates don’t change locally either, except in remote areas. Parcel carrier rate tables are based on their physical locations, their volumes across all their customers, and their pricing strategy. In other words, we can glean very little about how to optimize our local network from parcel carrier rate tables, except to see if we can do it less expensively ourselves.

How to Continually Innovate Cont’d

Fortunately, there is an excellent tool available that allows us to model local network scenarios—vehicle routing software. Why? Since local last-mile delivery is accomplished via truckload stop-off delivery routes and these tools accurately develop real-world last-mile delivery routes, they are excellent for modeling or costing out local network scenarios. Further, they are elegant at helping us better understand our true local market cost-service drivers and therefore help us uncover data-driven innovation opportunities. These tools allow us to model any “what if” innovation scenario we can come up with. Once we develop our local market network strategy, we then use our traditional network modeling software to integrate the local with our regional and national network strategy. We begin locally and work backward.

Vehicle routing software will answer the full range of questions retailers are asking: How do we best leverage our store assets to meet competitive service levels? Where can we offer free next-day or same-day delivery without substantially eroding margins? Which stores should we retain? Which stores should we evaluate for adding a micro fulfillment center (in a backroom or on a mezzanine)? What new locations should we evaluate for micro-fulfillment centers? Do we have the order density for operating our own last-mile delivery operation, or should we stay with a parcel carrier? Should we pursue co-delivery options? How can we evaluate whether co-delivery partner volumes are complimentary? What is the cost of serving each store? What geographies should we serve from our regional DCs versus the local market? The list goes on.

Examples

We can quickly cost out what is required for us to run our own fleet for local delivery. We can find the best central location for a local fulfillment center—either a new location or in-store. We can develop optimal service territories, simulating different demand scenarios for peak seasons, un-forecast- ed demand surges, and the corresponding impact on fleet utilization and costs. We can download the routes into Excel and develop cost allocation methodologies to determine the expense of serving each area and store. We can combine demand with other retailers to test whether the business volumes are complimentary. The tool is limited only by our imaginations.

Further, by analyzing the routes in a handful of representative local markets, the data itself will tell you how to innovate all markets for your business. For example, in the early days at Starbucks Coffee, we used vehicle routing software to develop our local network strategy. All Starbucks business segments—retail coffee, restaurant supply, and consumer direct—are eCommerce size shipments (UPS, FedEx), so the example fits today. After modeling a handful of representative local markets, the routes told us that, when a market reaches approximately 250 stores,6 we can operate a local DC and fleet at less expensively than using a parcel carrier. And when a market reaches approximately 500 stores, we could operate a roasting plant as well. From here, we used traditional network modeling software to design the best national strategy to serve each of our local market networks. This analysis saved us $44 million over 5 years. Route modeling will also help reveal significant innovation opportunities for your local network strategy.

A New Era of Network Tools

You may be wondering if there is an easier and faster way than using two tools and going through all these iterations? I’m glad you asked! Yes. Let’s first explain how these traditional tools were designed. Traditional network modeling and routing tools have been around for more than three decades, and the basic modeling approach has not changed. Both network and routing problems are too complex for true mathematical optimization. Network tools flow products from source to destination through the least expensive path of DCs and transport modes. It’s not rocket science, but their value is in quickly running a high number of scenarios.

Routing software is much more complex as it develops actual routes considering a larger number of factors: geography, traffic conditions, trailer capacity (cube and weight), trailer type, travel, stop and shift time, labor and mileage costs, etc. Thus, it uses heuristics to find the best or lowest-cost routes. It first assigns stops to vehicles and then keeps moving stops around— changing their sequence on trucks and moving stops among trucks—until it can no longer find a lower-cost set of routes. Furthermore, routing software has data visualization or the ability to graphically display the network and routes, which is very helpful in reviewing data. This is how today’s tools have been designed; network and routing are separate modules, and you will need to use the approach outlined above to model your network.

New Tool Advantages

Traditionally, network models are used to optimize the number and location of national and regional distribution centers, and routing models are used to optimize fleet operations for store and customer delivery. In today's rapidly changing local omni-channel markets, these models need to be integrated, and automated for rapid, continuous automation.

The advantage of Asher Supply Chain Designer over older packages—they developed the math to combine network and route optimization into one model. In other words, they can accurately model outbound last-mile route costs in addition to the point-to-point costs (TL, LTL, and parcel carriers). Why does this matter? The math to combine network and route costs is not simple. The sheer number of iterations one has to go through by modeling local routes and then integrating the local into the national network is cumbersome and time-consuming. The data needs to be exported, manipulated with another tool, and imported for each local market scenario. With a single tool, this is all eliminated. Your time to market for innovative scenarios and the labor cost to run the model is greatly reduced.

Supply Chain Designer is built on an AI platform to allow end-to-end process automation, which is essential for rapid, continuous operation to meet today’s rapidly changing world.