Distribution truck on the highway for last-mile urban deliveries

Definition

Last mile is the final segment of the distribution chain in which goods move from a hub, store, or consolidation center to the agreed delivery point with the customer or establishment, including attempts, windows, and proof of delivery.

Your Mexico City–Monterrey linehaul is weight-optimized. Then the south-zone delivery is rescheduled twice, the customer is not home, and cost per order doubles. Last mile does not forgive the spreadsheet that only watches rate per km.

If you run retail, 3PL, or e-commerce in Mexico, last mile is where margin and NPS fight on the same day. A TMS with routes and exception agents keeps chat from becoming your control tower, see TMS vs agents.

What last mile means in logistics

It is not only “the van at the end”. It includes route assignment, delivery window, attempt, retry, partial return, and POD. In B2B it may be delivery to the customer store/DC; in B2C, the home or pickup point.

Unit cost is high because the vehicle drops little volume per stop and every failed window implies unproductive km and hours.

Cost % and KPIs you should watch

In dense e-commerce, last mile often represents roughly 40–55% of total order logistics cost (varies by density and failure %). In B2B store distribution, relative weight falls but OTIF impact rises.

Do not optimize fuel alone. Optimize the attempt-and-window system.

Cost per delivery

What it measures: MXN / successful drop

MX warning signal: Rises with retries

Lever: Densify + windows

First-attempt success

What it measures: % deliveries on first visit

MX warning signal: <85% urban often fragile

Lever: Appointments / pickup points

Density

What it measures: Drops / hour or / km

MX warning signal: Long lightly loaded routes

Lever: Geographic clustering

OTIF / window

What it measures: On time and in full

MX warning signal: ±2 h windows breaking

Lever: Track & alerts

Digital POD

What it measures: % with evidence <X h

MX warning signal: Loose chat photos

Lever: ePOD linked to order

How to improve last mile in 5 steps

Prioritize density and first attempt; an electric vehicle will not fix a badly promised window.

Select a step to see detail

Step detail · 01

Measure an honest baseline

Step 1

Cost/delivery, first-attempt %, retries, OTIF by zone. Separate linehaul from last mile or you will lie to your margin.

4 last-mile myths in Mexico

If you chase these myths, cost per delivery will not fall.

1.More vehicles fix lateness

Without density or windows, more fleet only adds fixed cost. First fix the wave plan and first-attempt %.

Capacity ≠ control

2.The customer always wants same-day

Many accept reliable windows. Poorly executed same-day destroys margin and trust.

Reliable > heroic

3.WhatsApp is enough for the tower

It scales until it does not. Without a system, there is no defensible OTIF or retry audit.

Channel ≠ TMS

4.Last mile is only a B2C problem

Store deliveries with rigid appointments are also B2B last mile, with OTIF penalties included.

Mexican retail is demanding

Last mile with TMS and agents

The TMS coordinates orders and routes; agents watch exceptions and help tender capacity when a unit is missing.

Do not automate three fronts at once: if pain is service/OTIF, start with track & trace; if it is spot carrier cost, tendering; if it is invoice leakage from courier/3PL, audit, guide in TMS vs agents.

Key takeaways5 points
  1. Last mile is the final leg to the customer/store; it often concentrates 40–55% of logistics cost in dense e-commerce.
  2. KPIs that matter: cost per delivery, first-attempt %, densification, OTIF/window, and digital POD.
  3. In Mexico, traffic, security, and retail appointments change the equation: do not copy EU benchmarks without context.
  4. WhatsApp as a control tower does not scale: you need routes, exceptions, and evidence in a system.
  5. Improving last mile without aligned tendering/track/audit only moves pain to another link.

Is your last mile earning margin… or eating it?

Book a demo: windows, exceptions, and evidence to cut cost per delivery without breaking OTIF.
Book a demo

Frequently asked questions