SA · Logistics · Last-Mile Platform

Last-mile
delivery
re-engineered.

Clipboards. WhatsApp. Carbon-copy waybills.
Replaced with one platform in 6 weeks.

340+ Daily parcels
28 Drivers
4 Depots
83% 1st-attempt rate
83% delivery success
38 min → 90 sec booking
34% fuel saving
Dispatch Console Waybills Reports
All systems live
0 Delivered
0 In Transit
0 Failed
28 Drivers live
14:32 Avg ETA
Daily progress 0%
Live Map — Greater Johannesburg
Active At risk Late
D
JHB Central
D
Pretoria
D
Midrand
D
East Rand
K. Mahlangu
Parcels14 / 18 ETA depot16:42 StatusOn track
Driver Status 28 active
Live Events streaming
GPS sync · 30s · all 28 drivers
The Problem

A courier operation
running on paper.

47%
failed first-attempt delivery rate

No visibility.
No proof of attempt.

Drivers carried printed manifests and returned to the depot with hand-annotated outcomes. There was no live view of where any driver was, which parcels were delivered, or why attempts failed. Disputes with senders had no resolution mechanism — just a driver's word on a carbon-copy form.

Customer: "Where is my parcel?" — Operations: "Let me call the driver."
Daily Manifest — Paper-based process
WaybillRecipientStatus
FW-00841J. Dlamini✓ Del
FW-00842T. Nkosi✗ NR
FW-00843M. van Wyk✗ NR
FW-00844S. Patel? unknown
FW-00845A. Botha✗ WA
NR = No Response · WA = Wrong Address · handwritten
No timestamp. No photo. No GPS. No audit trail.
Waybill generation — manual process
01
Sender calls depot
~15 min
02
Operator types waybill
~8 min
03
Print + attach
~5 min
04
Manual capture to Excel
~10 min
Total per booking ~38 minutes · error-prone
38min
per waybill booking
340+
parcels per day

Manual waybills.
Manual everything.

Every booking was a phone call, a typed form, a printed label, and a duplicate Excel entry. With 340+ parcels a day, operations staff spent the majority of their shift doing data capture — not managing the operation. Errors were frequent: wrong addresses, duplicate waybill numbers, misrouted parcels.

On peak days, the booking queue backed up 2–3 hours. Parcels missed same-day cutoff. Senders escalated. Staff overtime spiralled.
0
real-time route optimisation

28 drivers.
No route intelligence.

Routes were assigned by the dispatcher from memory and experience. No system considered live traffic, parcel weight/volume, delivery windows, or depot return times. Drivers regularly ran inefficient loops, missed time-sensitive deliveries, and returned to the depot for re-routing — burning fuel and time.

Average fuel cost per driver per month was 34% higher than industry benchmark for equivalent parcel volumes.
Driver Route Efficiency — Before
Depot End backtrack
Route assigned from memory. No optimization. +34% fuel vs benchmark.
The Solution

Built in 6 weeks.
Live on day 43.

A full-stack web application: dispatcher dashboard, driver mobile PWA, customer tracking portal, and automated waybill engine — all in one deployment.

Automated Waybill Engine
01
Online booking form
~90 sec
02
Auto waybill + barcode
<2 sec
03
Auto-assign depot + route
<1 sec
04
SMS + tracking link sent
<5 sec
Total per booking ~90 seconds · zero errors
01

Waybill engine.
90 seconds. Automated.

  • Online booking form replaces all phone-in and operator typing
  • Barcode-labelled waybill generated and emailed instantly
  • Auto-assigned to correct depot based on pickup suburb
  • Sender and recipient both receive SMS with live tracking link
02

Dispatcher dashboard.
Live. Complete.

  • All 28 drivers on a live map — GPS updated every 30 seconds
  • Parcel status per driver: collected, in-transit, delivered, failed
  • One-click reallocation — move parcels between drivers in real time
  • End-of-day digital manifest auto-generated, no manual capture
Dispatcher Dashboard — Driver Overview Live
DriverParcelsDelETA depot
K. Mahlangu1411 ✓16:30
B. Sithole189 ⚠17:10
T. Fredericks224 ✗18:40 !
S. Khumalo1616 ✓Done ✓
GPS sync: 28 drivers active
Driver PWA — Mobile delivery workflow
FW-00843
M. van Wyk
12 Elm Street, Fourways, 2191
Delivered — capture photo
Failed attempt — select reason
No response at door
Wrong address
Refused delivery
POD photo + GPS coords attached automatically
03

Driver PWA.
Every delivery verified.

  • Progressive Web App — no install, works on any Android device
  • Scan barcode on collection — parcel status updated live
  • Successful delivery: POD photo + GPS coordinates captured
  • Failed attempt: structured reason code + auto-rescheduling triggered
04

Customer tracking.
Self-service. Always live.

  • Unique tracking URL sent by SMS at booking — no login required
  • Live status: booked → collected → in transit → out for delivery → delivered
  • ETA calculated from driver GPS + remaining stops
  • Failed delivery? Customer selects redelivery slot online
track.fastway.co.za/FW-00843
Waybill FW-00843 · M. van Wyk
Booked07:14
Collected09:22
Out for deliveryNow · ETA 14:45
Delivered
Results · 90 Days Post-Launch

Every metric moved.
Dramatically.

0%
First-attempt delivery rate
↑ from 47%
90 sec
Waybill booking time
↓ from 38 minutes
0%
Fuel cost reduction
↓ route optimisation
0%
POD coverage
↑ from zero
0 calls
Daily "where is my parcel" inbound calls
↓ from 80–120/day · all self-served via tracking portal
0
Build to live
Full platform, 4 depots
DimensionBeforeAfter WCG
Waybill booking38 min, phone call90 sec, online
Driver visibilityNone — call driverLive GPS, 30s refresh
Delivery proofCarbon copyPhoto + GPS + timestamp
Failed attemptsHandwritten noteReason code + auto-reschedule
Customer trackingPhone operationsSelf-serve portal
Route planningDispatcher memoryAlgorithm-optimised
End-of-day reportManual Excel captureAuto-generated
Dispute resolutionDriver's wordPhoto + GPS evidence
Built With

Stack overview.

Frontend
ReactPWA / Service WorkersLeaflet.js mapsRecharts
Backend
Node.jsExpressREST APIWebSockets (GPS)
Data
PostgreSQLRedis (live state)S3 (POD photos)
Integrations
Twilio SMSGoogle Maps APIZPL barcode printing
"
Before WebconstructGlobal, my operations manager spent three hours every morning on the phone — tracking drivers, answering "where is my parcel" calls, resolving disputes we couldn't prove either way. Now he manages the operation from one screen. The platform paid for itself in the first month.
RL
Riaan Louw Regional Operations Manager · Fastway Couriers SA
← Back to all case studies
Platform Capabilities

Everything you need to run
an intelligent business.

From fragmented systems to a unified decision-making engine.

Data Layer

Unified Data Layer

All your operational data — one schema, one source of truth, live across every site.

ERP
CRM
IoT
FM
HR
logo
Intelligence

Real-Time Intelligence

Live KPIs updating as events are captured on site.

91%
SLA Compliance
R1.74M
Capex EFC
Dashboards

Operational Dashboards

Group-level visibility across every site, contractor, and risk.

Site Overview Live
Centurion✓ 100%
Menlyn Park⚠ 94%
Eastgate SC✓ 100%
Sandton City⚡ Breach
Automation

Workflow Automation

Trigger → route → resolve. Zero manual intervention.

Trigger
Route
Resolve
SLA breach → FM manager notified
VO request → digital approval queued
Cert expiry → 7-day alert triggered
Predictive

Predictive Insights

Surface risk before it materialises. EFC modelling on every project.

Now Forecast →
Actual Forecast
Deployment

Rapid Deployment

Live within one week. Zero infrastructure required — cloud-native, mobile-first.

7days to live
0infra changes
100%cloud-native
Scope
Configure
Deploy
Live
09:14 Data schema validated across 14 sites
09:22 Contractor scorecard templates applied
09:31 Deploying BOQ library — 340 line items...
--:-- Going live...
Get Started

Live within one week. No commitment required.