Key takeaways
- Last-mile delivery consumes an outsized share of logistics spend — global studies consistently put it at ~50% of total shipping cost, and Indian operations are no exception.
- India’s national logistics cost has historically run high relative to developed markets (government estimates have ranged around 8–14% of GDP depending on methodology), which is exactly why per-order efficiency is a competitive weapon here.
- The two silent killers of cost per order are failed deliveries (each reattempt re-incurs most of the delivery’s marginal cost) and low drop density (few orders per route-km).
- The single most important metric to manage is cost per successful delivery, not cost per attempt — a ₹40 delivery with a 15% failure rate is really a ₹46+ delivery before recovery costs.
- Route optimization, first-attempt-rate improvement and fleet-mix rebalancing are the three levers with the fastest payback — typically visible within one quarter.
India last-mile cost benchmarks by segment (2026)
Directional ranges compiled from industry reporting, carrier rate cards and operator inputs. Your number depends on drop density, fleet mix, and service promise — compute it with the formula in the next section.
| Segment | Typical cost per delivered order | Primary cost drivers |
|---|---|---|
| Intracity e-commerce parcel (metro, high density) | ₹35–₹60 | Drop density, rider utilisation |
| E-commerce parcel, Tier 2/3 & upcountry lanes | ₹50–₹90 | Line-haul allocation, low density, RTO risk |
| Hyperlocal / quick commerce (per drop) | ₹25–₹50 | Batching rate, rider idle time, dark-store spread |
| FMCG / distribution beat delivery (per drop, van) | ₹40–₹80 | Stops per beat, vehicle utilisation |
| Pharma / cold chain | ₹90–₹200+ | Reefer assets, compliance, low batching |
| Furniture / appliances (big & bulky, 2-person) | ₹150–₹400+ | Crew cost, installation time, slot adherence |
Two structural notes on why these numbers behave the way they do:
- Density beats distance. A 60-stop metro route at 1.2 km between drops beats a 25-stop suburban route at 4 km between drops on cost per order, even if total kilometres are similar. Cost per order is mostly a density problem.
- Failure re-prices everything. At a 90% first-attempt delivery rate, 1 in 10 orders pays for two trips. Formally: effective cost per delivered order ≈ cost per attempt ÷ first-attempt success rate (plus recovery overhead). This is why failed delivery management is a finance topic, not just an ops topic.
How to calculate your true cost per delivery (the formula)
Most teams undercount. The complete formula:
Cost per delivered order = (Fleet fixed costs + Variable running costs + Labour + Overheads + Failure costs) ÷ Successful deliveries
Where, per month:
- Fleet fixed: vehicle EMI/lease/depreciation, insurance, permits, maintenance reserve
- Variable running: fuel (or energy), tolls, tyres/consumables per km × km run
- Labour: driver/rider salaries + incentives + helper/crew costs, including idle time
- Overheads (allocated): dispatch/ops team, hub handling for last-mile, software, telecom
- Failure costs: reattempt trips, RTO shipping, customer-support handling, refunds/appeasements
- 3PL component: for outsourced volume, the per-shipment invoice — plus your internal cost of managing exceptions on that volume (rarely zero)
Worked example — a 20-rider metro D2C fleet:
| Line | Monthly ₹ |
|---|---|
| Rider cost (20 × ₹22,000) | 4,40,000 |
| Vehicle + fuel (20 × ₹9,500) | 1,90,000 |
| Ops/dispatch team + software | 1,20,000 |
| Failure & RTO costs | 55,000 |
| Total | 8,05,000 |
| Delivered orders (20 riders × 45/day × 26 days × 92% success) | ~21,500 |
| Cost per delivered order | ~₹37.4 |
Run this once a month, per city, per fleet type. The segments where you’re above benchmark are your roadmap.
7 levers that reduce cost per delivery (ranked by speed of payback)
1. Route optimization (payback: weeks)
Fleet-wide route optimization typically cuts distance 10–20% and lifts orders per vehicle per day 15–30%. Because it attacks both the numerator (running cost) and denominator (orders delivered), it moves cost per order faster than any other lever.
2. Lift first-attempt delivery rate (payback: weeks)
Every point of first-attempt success removes a re-trip. The toolkit: accurate day-of ETAs with live tracking links, pre-delivery WhatsApp confirmation on COD, evening-window sequencing for residential stops, and NDR outreach that converts reattempts into scheduled (routable) events.
3. Rebalance fleet mix (payback: 1–2 quarters)
Match vehicle to drop profile: two-wheelers for dense small-parcel zones, Ace-class vans for bulky/consolidated beats, 3PL flex capacity for demand spikes instead of owning peak. Most fleets discover 10–20% of their runs are on the wrong vehicle class.
4. Increase drop density before adding vehicles
Tighten delivery-day promises by zone (serve each micro-zone on fewer days with fuller routes), consolidate multi-order customers, and use PUDO points for low-density pin codes. Density improvements compound with lever #1.
5. Automate dispatch (payback: 1 quarter)
A dispatch board that auto-assigns orders to riders/3PLs removes planner hours, cuts allocation errors, and reacts to exceptions in minutes instead of phone-call chains. Labour is usually the largest single line in the cost stack — planner productivity is part of it.
6. Instrument and pay for performance
Per-rider and per-3PL scorecards (cost per drop, FADR, SLA adherence, fake-NDR rate) turn averages into managed distributions. The gap between your best and worst rider on orders/day is often 2× — closing half of it is free capacity.
7. Cut empty and idle kilometres
Telematics + plan-vs-actual analysis exposes route deviation, personal-use km and hub dwell time. Even 5% km leakage on a mid-size fleet funds the software that finds it.
What good looks like: the metrics dashboard
Track these five weekly, per city and fleet type: cost per delivered order, first-attempt delivery rate (FADR), orders per vehicle per day, km per drop, and failure cost as % of delivery cost. If you can’t produce these numbers in under five minutes, that measurement gap — not fuel prices — is your biggest cost problem, and it’s what a delivery management system exists to close.
Frequently asked questions
How much does last-mile delivery cost per order in India? Typically ₹35–₹60 for dense metro e-commerce parcels, ₹50–₹90 for Tier 2/3 lanes, ₹25–₹50 per drop for well-batched hyperlocal, and ₹120–₹400+ for cold-chain or big-and-bulky deliveries. The spread is driven mainly by drop density, fleet mix and failure rates.
Why is last-mile delivery so expensive? It combines the smallest drop sizes with the highest stop counts and the most customer-driven variability (time windows, availability, address quality). Industry analyses consistently attribute about half of total shipping cost to the last mile.
How do I calculate cost per delivery? Divide total monthly last-mile cost (fleet fixed + variable running + labour + allocated overheads + failure/RTO costs) by successful deliveries in the month — using attempts instead of successes understates your true cost.
What is a good first-attempt delivery rate? Above 92% is healthy for Indian e-commerce parcels; best-in-class operations with strong ETA communication and NDR workflows sustain 95%+. Every percentage point directly reduces effective cost per order.
What reduces delivery costs fastest? Route optimization and first-attempt-rate improvement — both typically show measurable savings within weeks because they need no fleet changes, only better planning and communication.
Want to see ZenDMS on your operation?
Talk to our team for a 30-minute working demo, on your data, your lanes, your constraints. Schedule it here.