This post is about the four cost drivers, the density and routing economics, and the urban constraints, tested in the eyes of Amazon and Twiga Foods.
Table of Contents
Logistics Cost Drivers
Definition
Logistics cost drivers are the variables that determine the total cost of moving and storing goods. Alan Rushton, Phil Croucher, and Peter Baker, in The Handbook of Logistics and Distribution Management (now in its sixth edition), identify four primary drivers: distance, density, speed, and variability. Each driver interacts with the others, and managing them together is the core task of logistics design. Martin Christopher, in Logistics and Supply Chain Management, frames the same drivers as the levers that connect logistics strategy to customer service.
The four cost drivers are:
- Distance — the physical length of the movement; cost scales with kilometres travelled
- Density — the concentration of deliveries within a geographic area; higher density reduces cost per delivery
- Speed — the required delivery window; faster delivery increases cost
- Variability — the unpredictability of demand, routes, or volumes; variability increases cost
I want you to notice why these four matter together, because the standard treatment of logistics presents them as separate levers. They are not. Distance and speed are obvious cost multipliers, but density is the one that does the most work in urban delivery. Delivering fifty parcels on a single street costs far less per parcel than delivering fifty parcels spread across fifty streets. Variability adds cost because firms must hold capacity to handle peaks that occur infrequently. The practical implication is that firms can reduce last-mile cost by increasing delivery density through route consolidation, hub networks, or pickup points; by relaxing speed requirements through scheduled delivery windows; and by reducing variability through demand forecasting and scheduling.
Firms can act on five primary levers to reduce last-mile cost:
- Consolidation — combining deliveries to increase density per stop
- Scheduled windows — reducing speed requirements by offering broader delivery windows
- Pickup points — shifting delivery to lockers or partner locations to reduce per-address cost
- Demand forecasting — reducing variability through better prediction of order volume and location
- Route design — reducing distance by optimising delivery sequence
I will put it this way. Last-mile cost is not a fixed number. It is determined by design choices. Firms that treat last-mile delivery as a design problem rather than just an operational problem find material cost reductions that operational efficiency alone cannot achieve.
Delivery Density and Route Economics
Definition
Delivery density is the number of deliveries per unit of geographic area served. Route economics is the analysis of how route design, vehicle selection, and delivery sequence determine cost per delivery. Density and routing are interdependent. Higher density enables more efficient routes, and better routing enables firms to serve more customers per vehicle per day. The operations research literature, including the vehicle routing problem tradition that runs from Dantzig and Ramser (1959) through to contemporary algorithmic solutions, treats this as the central optimisation problem in last-mile logistics.
Density and routing economics rest on four measures:
- Delivery density — deliveries per square kilometre or per route
- Stops per hour — the operational metric that determines driver productivity
- Cost per delivery — the total cost of a route divided by deliveries completed
- Vehicle utilisation — the proportion of vehicle capacity actually used per route
I will not pretend the density argument is subtle, because it is not. Delivery density is the single most powerful lever in last-mile economics. Doubling the number of deliveries per square kilometre does not double cost; it substantially reduces cost per delivery because fixed route costs, including driver time, vehicle cost, and fuel, are spread over more stops. Firms increase density by concentrating demand through marketing, geographic targeting, or pickup points; by designing hub networks that stage parcels closer to delivery zones; and by consolidating deliveries into scheduled windows that group nearby orders together. Route optimisation software can improve density by ten to twenty per cent without any change to physical infrastructure.
The practices that improve density most reliably are:
- Hub placement — positioning hubs closer to delivery zones to reduce line-haul distance
- Zone grouping — clustering orders by delivery zone to increase stops per route
- Scheduled delivery — offering fixed windows that concentrate demand into predictable time slots
- Pickup point integration — shifting a portion of deliveries to lockers or partner locations
- Vehicle selection — matching vehicle size to route density and delivery type
I want you to see that density is a self-reinforcing advantage. Firms with high density have lower cost per delivery, which allows them to price aggressively, which attracts more demand, which increases density further. That dynamic is why last-mile logistics markets tend toward concentration, and why the largest operators in any given market maintain a structural advantage over smaller competitors who cannot match their density.
Routing Technology and Urban Logistics
Definition
Routing and delivery technology include the software, hardware, and data systems that determine how deliveries are planned, sequenced, and executed. Urban logistics refers to the specific challenges of delivering in dense urban environments, where traffic, parking, address accuracy, and regulatory constraints all affect cost and service. The OECD’s International Transport Forum and the MIT Center for Transportation and Logistics have both documented the growing regulatory pressure on urban freight, from congestion charges and low-emission zones to delivery time windows and vehicle-size restrictions.
Routing and delivery technology rest on four elements:
- Route optimisation software — algorithms that determine the sequence and timing of deliveries
- Telematics and tracking — GPS and sensor systems that monitor vehicle and driver performance
- Address intelligence — data systems that resolve incomplete or ambiguous delivery addresses
- Urban constraints — traffic congestion, parking limitations, delivery time windows, environmental regulations
I want you to see why technology matters more in last-mile than in any other part of the supply chain. Routing technology produces cost savings that are difficult to achieve through operational discipline alone. Optimisation algorithms can reduce total distance by ten to fifteen per cent compared with manual routing, and telematics can identify driver behaviour issues that increase fuel cost. In urban environments, additional constraints emerge. Congestion extends travel time unpredictably. Parking restrictions delay deliveries. Clean-air zones restrict vehicle types. Cities in Europe and Asia have increasingly imposed urban logistics regulations that require firms to redesign routes and vehicle fleets. In emerging markets, address accuracy is often the binding constraint. Many customers live in areas without formal addresses, and delivery requires local knowledge and phone-based coordination.
The technology and compliance levers that matter most are:
- Algorithmic routing — software-driven sequencing to minimise distance and time
- Real-time re-routing — dynamic adjustment when traffic or demand shifts
- Telematics — monitoring vehicle performance, driver behaviour, and route adherence
- Address verification — geocoding and confirmation to reduce failed deliveries
- Urban compliance — vehicle selection, time windows, and route design to comply with local regulations
My reading is that urban logistics constraints are becoming regulatory rather than physical. Firms that anticipate regulatory changes and redesign their networks in advance gain cost advantages over competitors that react to change. That is the shift that has happened over the last decade, and it will continue as cities tighten emissions and congestion rules.
Last-Mile Logistics in Emerging Markets
Definition
Emerging-market last-mile delivery refers to the specific constraints and opportunities that shape delivery in markets where road infrastructure, address systems, payment methods, and consumer expectations differ from developed markets. These conditions require delivery models that are adapted rather than imported. The World Bank’s logistics performance research and the African Development Bank’s trade and transport reports document the same pattern: emerging markets offer leapfrog opportunities alongside structural constraints that developed-market playbooks do not address.
Emerging-market last-mile delivery rests on five constraints and opportunities:
- Infrastructure constraints — road quality, traffic density, and vehicle access limitations
- Address informality — reliance on landmarks, phone contact, and local knowledge rather than formal addresses
- Payment methods — prevalence of cash-on-delivery and mobile money rather than card payment
- Consumer expectations — different tolerance for delivery windows and communication channels
- Cost structures — lower labour costs offsetting higher per-delivery complexity
I will not pretend the emerging-market case is simply a variation on the developed-market playbook, because it is not. Emerging-market last-mile delivery operates under constraints that developed-market models do not address. Address informality forces firms to build verification and coordination processes that would be unnecessary in markets with formal postal systems. Cash-on-delivery prevalence requires driver cash handling and reconciliation. Low delivery density in some areas makes urban-style consolidation uneconomic. Successful emerging-market delivery models combine technology with local knowledge: mobile phone coordination, agent networks, and pickup points where formal addresses do not exist. M-PESA and similar mobile money systems provide payment infrastructure that substitutes for cards and enables new delivery models.
The mechanisms that work most reliably in emerging markets are:
- Agent-assisted delivery — using local agents as delivery points where formal addresses are absent
- Mobile coordination — using phone and SMS to confirm deliveries in real time
- Mobile money integration — accepting payment through M-PESA and similar systems
- Hub-and-spoke with local agents — combining urban hubs with last-leg delivery by local operators
- Cash handling controls — processes to reconcile cash-on-delivery without security incidents
My reading is that emerging-market last-mile delivery is not a diminished version of developed-market delivery. It is a different model, adapted to different constraints, and in some respects more innovative than developed-market approaches. Firms that treat it as a lower-cost version of the same game misread the market.
Case Study
The two clearest public illustrations of last-mile economics are Amazon, whose logistics network is the largest single investment in delivery infrastructure by a private company, and Twiga Foods, the Kenyan food distribution company that built a last-mile network for fresh produce from rural farmers to urban vendors. They sit at opposite ends of the resource spectrum and operate under very different market conditions, and reading them together shows what is transferable across contexts and what is not.
Amazon delivered more than six billion packages globally in 2025. Its network is designed for density. Delivery stations are placed within short distances of high-density urban zones. Routing software sequences deliveries for maximum stops per hour. Pickup point networks reduce per-address delivery cost in apartment-dense areas. The company’s same-day and next-day programmes are economically viable only because of this density, and the density advantage compounds with each additional delivery. Twiga Foods operates in a market where addresses are often informal and delivery instructions rely on phone calls and local landmarks. The company uses mobile technology to coordinate with thousands of small urban vendors, aggregates orders into consolidated routes, and stages produce at urban hubs before final delivery. Its economics depend on achieving density in Nairobi’s dense urban markets, where per-delivery cost can support the business model. Twiga has demonstrated that last-mile delivery in emerging markets can be profitable when adapted to local conditions rather than imported from developed-market templates.
I want you to see what the two cases establish together. In both, last-mile delivery was designed rather than inherited. Amazon built the delivery station network and routing technology from scratch. Twiga built the agent coordination model and hub-and-spoke system from scratch. In both, the design was tuned to the density and constraint profile of the market it served. Amazon’s network would not work in Nairobi, and Twiga’s model would not work in Manhattan. In both, the cost advantage came from density achieved through design, not from squeezing more efficiency out of an existing operation. The cases also show what the frameworks do not provide. Neither firm achieved its delivery economics by adopting a template. Both built the model from the ground up and adjusted it as the market revealed what mattered. The frameworks describe the cost drivers and the design levers. The discipline is what makes them work.
Conclusion
The Amazon and Twiga Foods cases together establish a proposition that is more useful than either case alone: last-mile logistics is a design problem, not an efficiency problem. Firms that treat it as efficiency try to squeeze cost out of an existing operation and hit a floor. Firms that treat it as design rebuild the network around the density and constraint profile of their market and find cost reductions that were not available through operational improvements.
My detailed conclusion is this. The four cost drivers from Rushton, Croucher, and Baker — distance, density, speed, and variability — are the correct analytical frame, but the standard treatment of them undersells density. Density is the lever that compounds. Distance and speed are linear multipliers. Variability is a cost of uncertainty. Density is the only one of the four that generates a self-reinforcing advantage, because higher density lowers cost per delivery, which allows lower prices, which attracts more demand, which further increases density. Firms that ignore this dynamic and focus instead on squeezing fuel cost or driver productivity are optimising the wrong variable. The technology story reinforces the point. Routing software, telematics, and address intelligence produce real cost reductions of ten to fifteen per cent. But the biggest gains in last-mile economics come from network design decisions that routing software can only optimise within, not replace. The frameworks also understate how different emerging-market delivery is from developed-market delivery. The standard playbook assumes formal addresses, card payment, and dense urban demand. None of those assumptions hold in most of sub-Saharan Africa and much of South Asia and Latin America. Firms that import developed-market delivery models into emerging markets struggle. Firms that build delivery models around the density, address, and payment realities of the market they serve succeed. That is the most important finding in the entire last-mile literature, and the standard textbooks still treat it as a footnote.
My recommendations follow from that conclusion. First, treat last-mile delivery as a network design problem before treating it as an operations problem; the biggest cost reductions come from hub placement, zone design, and delivery model choices, not from squeezing the existing operation. Second, measure density explicitly and treat it as the primary KPI; if density is not improving, cost per delivery will not improve either. Third, use technology to optimise within the design, not to substitute for design; routing software is a multiplier, not a solution. Fourth, when operating in emerging markets, build the delivery model from the local constraints up rather than importing a developed-market template and adjusting it. Fifth, anticipate regulatory change in urban logistics and design for it; the firms that redesign their fleets and routes before regulations force them to will hold a cost advantage over competitors that react later.
The final argument of this post is that last-mile logistics is not what you optimise. It is what you design. The cost drivers describe the problem, and the design choices determine whether you solve it.
Read Also on Kat-Syd Resources Hub
Management Principles Series: Planning, Organizing, Staffing, Directing & Controlling — Foundational management functions that govern how firms plan, organise, and control their operations.
Functional Management — The Pillars — How core business functions including operations and logistics are organised and measured.
Entrepreneurship and Innovation Part I — Ideation and Opportunity Recognition — The entrepreneurial mindset, source of innovation, and opportunity assessment.
Written by Kateule Sydney — Researcher and Writer
© 2026 Kat-Syd Resources Hub. All rights reserved.
Kat-Syd Resources Hub — Educational case studies and analytical reference