First-Mile & Last-Mile

A mid-sized e-commerce retailer with operations in several urban and suburban regions was facing unpredictable delivery costs. While order volumes were growing steadily, the company lacked visibility into where inefficiencies were concentrated in its first-mile (warehouse to hub) and last-mile (hub to customer) operations.

Deliveries relied on a mixed fleet of third-party couriers, company-owned vans, and small EVs/e-bikes in urban centers. But the absence of clear cost breakdowns meant leadership couldn’t decide how to balance investments across these modes. Partnering with Sphere, the retailer deployed a cost analytics platform that mapped real cost drivers, compared delivery mode performance, and simulated scenarios for future network design.

Sphere built a First-Mile & Last-Mile Cost Analytics Engine that gave the retailer transparency into true delivery costs and a roadmap for optimizing its fleet mix.

End-to-End Data Integration

The first step was unifying all relevant operational data into a single platform. This included courier invoices by zone, van mileage and fuel logs, EV charging costs, e-bike usage, and warehouse handling times. By consolidating everything into one source of truth, the company gained full visibility into cost per order across first- and last-mile operations.

Cost Driver Mapping

Analytics then revealed how costs behaved for each delivery mode. Couriers were highly variable, especially in suburbs with lower order density. Vans carried higher fixed costs but performed best on bulk suburban routes. EVs proved stable on a per-mile basis and worked well in mid-range city deliveries. E-bikes had the lowest cost per package in dense urban areas but were inefficient beyond short-range delivery. This mapping provided leadership with its first evidence-based comparison.

Scenario-Based Simulations

The platform ran simulations to test different fleet strategies under real-world conditions. Scenarios included a courier-heavy model, a van-heavy suburban network, an EV/e-bike urban model, and a mixed allocation balancing all modes. Each was stress-tested against demand spikes, fuel price volatility, labor shortages, and seasonal delivery patterns, showing the cost trade-offs of each approach.

First-Mile Optimization Insights

The analysis also uncovered inefficiencies in the first mile. Roughly 15–20% of excess costs stemmed from vans and couriers waiting idle due to delayed loading. Recommendations included staggered dispatch windows, dynamic warehouse-to-hub handoffs, and tracking loading times as a KPI. These changes alone promised immediate savings without altering the delivery network.

Per-Order Cost Forecasting

With advanced forecasting models, the system achieved 95% accuracy in predicting per-order delivery costs. Finance teams could now model budgets and customer pricing strategies with confidence, avoiding surprises from fluctuating courier fees or fuel costs.

ROI Forecasting & Roadmap

Finally, Sphere’s engine projected the return on investment for different delivery strategies. It showed clear breakeven points for scaling EV fleets, pinpointed where e-bikes outperformed couriers in city centers, and confirmed that vans remained the most cost-efficient option for suburban clusters. Leadership now had a phased roadmap for future fleet investments based on hard data.

Sphere’s cost analytics engine gave the retailer end-to-end visibility into its logistics costs. By breaking down the economics of couriers, vans, EVs, and e-bikes, the company gained the clarity it needed to reduce waste, stabilize delivery expenses, and make smart investment decisions. With predictive models and scenario testing, leadership could confidently design a cost-efficient network aligned with both financial goals and sustainability targets.