“Middle mile” has less buzz than first mile and last mile logistics and is most often considered the part where not much of an optimization can happen. Much of the focus today is on last-mile deliveries and even first-mile services that link vendors and fulfillment centers. So, logistic professionals and distribution practitioners sometimes consider it as a most “boring” part of the supply chain as middle-mile logistics is relatively easy to manage and automate. But is it so?
Middle-mile delivery distances are different from sector to sector. Nevertheless, with the accelerated growth of e-commerce increasing requirements are creating burdens on the logistics chain, including middle-mile. As it is clear from the report represented by Deutsche Bank in 2019, networks are becoming more fragmented supporting the positioning of inventory near the final destination. For example, forced by Amazon’s “Prime” delivery service traditional retailers on average have cut their delivery times by as much as 30 percent in the past few years. We also observe an increase of fulfillment sites near densely populated demand centers. So a shortened and fragmented middle-mile creates uncertainties and retains higher volatility typical to the last mile.
What is “middle-mile” logistics for different sectors?
Lets first define what is the “middle mile”. Now let’s define what is the “middle mile logistics” for different industries?
Wholesale distributors are the middleman between producers and retailers with slight differences from the industry to industry. It is one of the more complex industries driven by high numbers of SKUs, customers, suppliers and transactions, and their pricing and rebate structures. They can be grouped in major ones such as food, beverages, pharmaceutical, electronics, industrial and building materials. According to the National Association of Wholesale Distributors (NAW) total wholesale distribution revenue ended the year at $5.970 trillion, just slightly below the April 2019 record high. GDP for the year totaled a record high $21.734 trillion with a growth rate of 4.0%. The Wholesale Distribution industry is 27.5% of U.S. GDP1. From the total supply chain perspective wholesale distribution is a combination of first and middle-mile deliveries. While a small distributor can work having only one distribution center serving a limited area, large distributors, may have hundreds of distribution centers in the majority of states and Canadian provinces (we have in mind only the domestic part of the supply chain without considering international movements here). We observe changes in the industry in the form of growing concentration as well as the emergence of disruptive business models. For example, Amazon’s sales platform is supported by more than 2 million third-party sellers worldwide. Of those sellers, 26% sell products using a wholesale sales model. So, changes drive growth for some distributors or create significant disruption in the value chain for others. Managing these shifts to ensure sustainable business performance is a priority for wholesale distributors, particularly as new competitors emerge and apply additional pressures.
Unlike the wholesale distribution, retail distribution can have distinct first, middle and last miles in their logistic supply chains. While smaller retailers do only the last mile, larger retailers with multiple distribution centers may have all three. So, the middle-mile deliveries for retail businesses are the part of the distribution from DCs to their stores. Retailers may use both their private fleets, 3rd party carriers or the combination of both.
According to the latest report by PwC and the National Retail Federation (NRF) – released in May 2020 – the industry’s total GDP impact was $3.9 trillion, accounting for 18.7 percent of US GDP in 2018. The supermarket & Grocery stores industry grew at a rate of 1%, reaching $654.6 billion in 2019. Speciality Stores such as Home Depot, Best Buy, etc. currently represent 11% of retail sales. According to the latest figures by USDA, grocery stores, including supermarkets and smaller grocery stores (except convenience stores) accounted for the largest share of store sales (92.2 percent), followed by convenience stores without gasoline (4.5 percent). Specialized food stores, including meat and seafood markets, produce markets, retail bakeries, and candy and nut stores, accounted for the remaining 3.3 percent of the total. According to NRF latest data3, the main retailers in US are:
- Walmart, with a turnover of $ 387.66 billions in retail sales
- Amazon.com ($120.93 billion)
- The Kroger Co ($119.70 billion)
- Costco ($101.43 billion)
Transportation and logistics
Third party logistics providers, trucking companies and owner operators serve both retail and wholesale distribution industries. According to the U.S. transportation data the freight transported by private trucks is almost equal to the for-hire sector deliveries.
Many retailers and wholesalers outsource their logistics to 3PLs and 4PLs hence delegating the middle-mile logistics management function. Speed is becoming the main factor in organizing logistics from ports or production sites to distribution centers. Unlike many retailers and wholesalers, 3PLs manage freight for multiple customers hence have wider options in consolidation tactics. They particularly can alternate between full truckload (TL) or Less than truckload (LTL) movements to fasten up the delivery process. While it’s obvious in the case of first and last-mile deliveries, middle-mile delivery optimization is less visible but not a less important and difficult task.
This is a movement of goods from large DC’s to the growing number of smaller fulfillment centers. It is where much of the work occurs that makes logistics run, chiefly, consolidations and de-consolidations. It encompasses a phalanx of large distribution centers used to shift products around the country. It is also where much of the delivery distance is covered. The flagman of middle-mile distribution operationations is apparently Amazon.com Inc. In the upcoming future Amazon plans to open 1,000 small delivery hubs in cities and suburbs all over the U.S., according to people familiar with the plans5. The facilities, which will eventually number about 1,500, will bring products closer to customers, making shopping online about as fast as a quick run to the store. Other large and small companies are also keeping up the pace so the growth of smaller regional fulfillment centers is expected to rise. According to research from commercial real estate firm CBRE Group, in 2019 rents for warehouses between 70,000 and 120,000 square feet rose by 33.7 percent in the past five years. The average price is now $6.67 per square foot. Availability for these smaller warehouses also shrunk from 11.3 percent to 7.4 percent in the same time frame. Despite their help to speed up deliveries, more fulfillment centers mean higher cost and more pressure on making them efficient.
Two big questions of the middle-mile logistics
Are lanes static?
The number one misconception about the middle-mile is that it is “boring” and there is nothing to optimize. As we already pointed out, changing supply chains add volatility and increases the need of applying different strategies and scenarios. Moreover, it is important to deploy efficient strategies and tactics both for long term supply chain and daily delivery planning. For example, middle-mile logistics might alternate between three “legs”:
- Shipment of high volume items directly to DCs
- Truckloads of multiple SKUs can be used to reallocate inventories between distribution centers
- Deliver multiple SKU loads to smaller footprint urban fulfillment centers
The volatile demand and tight delivery windows may force shippers to continuously find balance in the logistic network orchestration. There should be proper tools to do that having in mind multiple optimization goals and constraints.
Retail distributors with multiple DCs also face the same dilemma of efficient logistics planning. Decisions about DC location and configuration change faster than before as retailers should adapt to changing demand. Volatile demand also creates inefficiencies for day to day operational planning. Retailers can deploy the same routes as delivery volumes so the service times vary quite a lot. Some DCs cover quite large areas so there could be 2-3 days multi-stop trips. So retailers should either add smaller DCs closer to their stores or optimize the delivery process to get more work done with the same or fewer resources. As transportation costs can reach up to 50% of total logistics costs, route and vehicle planning can either assure or dry up profits in a low margin environment.
Why visibility matters?
When it comes to visibility, logistic managers usually imply to the consumer that they know where your truck is. While it is good information to have, especially in real time, locating your delivery is one (small) part of the total visibility. It should help to not only track deliveries but also make preventive operational management and long term data analytics and efficiency optimization.
Capturing data: Capturing data is essential as it is created in real time or near time and distributed across different operational decision makers. It particularly includes stops or order level updates of estimated times of arrivals (ETA) at a driver, customer, dispatcher or DC level. Post factum delivery data such as late deliveries and actual service times is also important to capture and store properly. To minimize late deliveries by implementing dynamic routing, dispatchers need timely data that shows ETAs with service times and other variables not being produced by GPS tracking software.
Aggregating data: Capturing and aggregating delivery data per different instances and layers is the key for efficient operational and strategic middle-mile network management. The difficulty of it is that systems across different parts of the logistics chain are captured and stored differently, so it is not always possible to have different levels of data aggregation without a synchronized ecosystem.
Supporting decision-making: Aggregated data is a foundation for retrospective and prospective data analytics. One of the major problems of growingly complex logistic supply chains is the absence of full digitization and synchronization of all the necessary information across different layers of an organization.
How Less Platform helps to plan and orchestrate middle-mile logistics
Unlike last-mile deliveries, middle-mile loads can take several days to deliver, although companies are usually trying to locate their DCs within a daily dispatch limit. It creates problems for automatization of the load building and route planning process as the majority of solutions in the market does only daily planning. Less Platform tuned its route planning and load building algorithms to factor in multi-day multi-stop routing.
The other important feature of Less Platform is the multiple depot or DC planning functionality. It is very convenient for the companies that have centralised operational planning activities. Secondly, it helps to generate data for separate DCs as well as aggregate it on a company level. As we have already mentioned, the number of DCs are increasing and moving closer to customers making both operational and strategic routing harder and potential inefficiencies bigger.
Two important constraints that make it hard to plan are daily delivery windows and variable service times. These two constraints can also drastically affect capacity used and routes planned. The absence of proper planning tools can make a planner work daunting and consume too much time. It is also the main cause why distributors use excessive capacities, drive unnecessary miles and have late deliveries. These effects are being amplified due to increased demand variability, speed of deliveries and tightening delivery windows.
Automatic routing easies the job but there are plenty more situations requiring manual planning. Less platform provides dynamic planning functional for both pre-dispatch and post-dispatch stages. After getting a route plan, a route planner can reroute different stops by assigning them to different routes both individually and in bulk. Then sequencing algorithms will re-optimize the total plan. After making changes algorithms assess feasibility of delivery windows by checking with recalculated ETAs. All the changes are instantly visible for drivers post-dispatch.
And finally, planned loads can be scheduled both daily and weekly to respective drivers. Less Platform’s dispatch management software is connected to the driver app so all drivers have planned loads once they are assigned to them.
After the dispatch, the driver app updates all ETAs and synchronizes them with the dispatch board. Less Platform’s algorithms then check potential late deliveries by comparing ETAs with delivery windows. This way dispatchers can be proactive to avoid late deliveries or dynamically optimize the delivery process otherwise.
Vardan Markosyan is the CEO at Less® Platform
MBA from the University of Chicago Booth School of Business
PhD in Economics from the Institute of Economy of NAS RA
He spent decades of research and consultancy on business process optimization and system design