Last Mile Delivery Optimization: 7 Proven Strategies

Why Last Mile Is Your Biggest Cost Problem

Last mile delivery — the final leg from distribution centre to recipient — represents up to 53% of total shipping cost according to Capgemini research. It is also the most complex: high stop density, unpredictable recipient availability, urban traffic, and time window commitments all compress into the hours between 08:00 and 18:00.

For Swiss logistics operators, these challenges are amplified by geography, linguistic diversity, and customer expectations shaped by e-commerce. Getting last mile right is not optional — it is the competitive differentiator.

Here are 7 strategies that consistently move the needle.

Strategy 1: Dynamic Routing Over Static Planning

Static route planning — building a fixed sequence the night before and running it regardless of what happens — was the only option before real-time data. Today it is a significant liability.

Dynamic routing recalculates in response to real-world events: traffic delays, failed first attempts, added stops, vehicle breakdowns. A single re-optimisation triggered by a 30-minute traffic jam at stop 8 can save 45 minutes of downstream cascade delays.

Fleet Planner runs dynamic re-optimisation continuously. Dispatchers can trigger manual replans in under 60 seconds, and the Autopilot mode handles routine replanning automatically without dispatcher intervention.

Impact: 10–15% reduction in route time through the shift.

Strategy 2: Electronic Proof of Delivery with Exception Capture

First-attempt delivery failure is the most expensive event in last mile logistics. The failed attempt costs as much as a successful one — driver time, vehicle cost, fuel — and triggers a redelivery cycle.

ePOD systems with structured exception capture change this by creating data. When a driver marks "no recipient present" with a GPS-stamped timestamp and mandatory photo, you gain:

• Evidence for customer dispute resolution
• Data on which addresses consistently fail first attempts
• Trigger for immediate customer notification to reschedule

Driver Pro captures every exception type with mandatory structured fields, feeding exception analytics back to dispatchers in real time.

Impact: 15–25% reduction in redelivery rate through proactive customer communication triggered by ePOD exceptions.

Strategy 3: Proactive Customer Notification Windows

The primary cause of failed first attempts is recipient absence. The primary cause of recipient absence is insufficient notice.

SMS or push notifications with a 2-hour delivery window — sent 90 minutes before the driver arrives — convert passive recipients into active ones who plan around the delivery. Companies that implement delivery window notifications see first-attempt success rates improve by 20–30 percentage points.

The key is tying the notification to real-time route progress, not a morning estimate. "Your delivery is 3 stops away, estimated arrival 14:35–14:55" is dramatically more effective than "Your delivery is scheduled for this afternoon."

Impact: 20–30% improvement in first-attempt delivery rate.

Strategy 4: Time Window Optimisation

Not all delivery time windows are created equal. Customers requesting "morning" slots at remote addresses while also requesting "morning" slots at urban stops create routing conflicts that increase total distance significantly.

Analysing your historical delivery data reveals which time window combinations are geometrically incompatible — forcing routes that double back unnecessarily. Offering customers a set of pre-validated time windows (rather than free-form selection) can reduce total daily route distance by 8–12%.

Impact: 8–12% reduction in total daily kilometres driven.

Strategy 5: Density Routing for Urban Operations

Urban delivery in Zürich, Geneva, or Basel has a different optimal strategy than rural delivery. In dense urban areas, the optimal approach is zone-based density routing: concentrate all deliveries in one zone before moving to the next, minimising transit between stops.

Most generic routing algorithms optimise for distance, not density, producing routes that bounce between zones inefficiently. AI optimisers that learn your specific urban geography — including pedestrian zones, loading restrictions by time of day, and parking availability by address — deliver significantly better results.

Fleet Planner's Swiss urban routing engine is pre-loaded with city-specific loading zone data and time-based access restrictions for all major Swiss cities.

Impact: 12–18% reduction in urban stop time through better zone concentration.

Strategy 6: Fleet Right-Sizing

Running 12 vehicles to cover work that optimally requires 9 is a common result of organic fleet growth. Each excess vehicle carries fixed costs: lease payment, insurance, maintenance, and the minimum viable driver shift.

Regular fleet right-sizing analysis — comparing actual vehicle utilisation (stops completed vs. capacity) against demand patterns — identifies excess capacity. In practice, most fleets running manual planning are 15–25% over-capacity because planners conservatively over-provision vehicles to protect against uncertainty.

With AI optimisation eliminating most of that uncertainty, right-sizing to actual demand becomes feasible. Reducing from 12 to 10 vehicles on a route where that is genuinely sufficient saves CHF 60,000–80,000 per year in fixed costs.

Impact: 15–25% reduction in total vehicle operating cost.

Strategy 7: Driver Performance Analytics

Individual driver efficiency varies significantly even on identical routes. A 15% difference in stops-per-hour between your best and average driver translates directly into route completion time and overtime.

Driver performance analytics — dwell time per stop, deviation from optimised route, exception rate, customer rating — identify both coaching opportunities and systemic route design issues. Often, "slow" drivers are revealing stops that genuinely require more time due to access difficulty, and the route design should be adjusted.

Driver Pro captures granular per-stop timing data automatically. The dispatcher dashboard aggregates this into performance trends by driver, route, and customer — enabling targeted coaching conversations backed by data rather than impressions.

Impact: 8–12% improvement in average stops-per-hour through targeted coaching and route adjustment.

Compounding the Gains

These strategies are not mutually exclusive — they compound. A company implementing all seven typically sees:

• 30–40% reduction in total last mile cost
• 25–35% improvement in first-attempt delivery rate
• 15–20% reduction in fleet size for equivalent volume
• Customer satisfaction improvement measurable in Net Promoter Score

The sequencing matters: start with dynamic routing and ePOD (Strategies 1 and 2) as they deliver the largest immediate ROI and create the data foundation for the remaining strategies. Add customer notifications (Strategy 3) next for the fastest customer-visible improvement. The remaining strategies can be implemented in parallel over the following quarter.