The short answerEvery morning began with phone calls, WhatsApp messages, and spreadsheet updates to assign routes, confirm driver availability, and communicate delivery schedules to clients.
The short answer
Every morning began with phone calls, WhatsApp messages, and spreadsheet updates to assign routes, confirm driver availability, and communicate delivery schedules to clients.
What Is The Challenge: Manual Dispatch Coordination Consuming the Week?
Direct answer: A mid-sized logistics company operating in South Africa's Western Cape and Gauteng corridors was losing roughly 22 hours per week to manual dispatch coordination. According to Deloitte's 2024 Global Supply Chain Survey, 67% of logistics companies still rely on manual processes for more than half of their dispatch operations. This firm was no exception.
Current as of 31 May 2026: This article has been reviewed for the 2026 South African AI, SEO, and automation market. Pricing, platform capabilities, Google rich-result rules, and AI model features change quickly, so verify live vendor documentation before procurement. For privacy and data handling, use the Protection of Personal Information Act as the baseline; for search and structured-data implementation, use Google Search Central.
Their dispatch team of six coordinators managed a fleet of 38 vehicles across two provinces. Every morning began with phone calls, WhatsApp messages, and spreadsheet updates to assign routes, confirm driver availability, and communicate delivery schedules to clients. By mid-morning, the first round of changes would arrive: cancelled deliveries, urgent additions, vehicle breakdowns, and client rescheduling requests.
Each change triggered a cascade of manual updates across multiple systems and communication channels. The team was working hard. But they were working hard on coordination, not on the strategic decisions that actually improve delivery performance.
TL;DR: An investigation-first SOP audit of a South African logistics firm's dispatch operations uncovered 9 manual touchpoints and 3 redundant approval loops. Phased AI implementation eliminated the redundancies and reduced manual coordination by 22 hours per week, while the error rate dropped from 12% to under 3%.
Note: This case study is illustrative, drawn from composite experiences across multiple logistics engagements. Specific figures represent realistic outcomes based on our work in the South African logistics sector.
For a broader view of AI in South African logistics, see our guide to AI-powered route optimisation.
What Did the Investigation Process Involve?
Before recommending any technology, the first step was mapping every standard operating procedure involved in the dispatch workflow. This investigation phase lasted two weeks and involved shadowing every member of the dispatch team.
Week 1: Process Shadowing
We embedded with the dispatch team during a typical working week. Not interviewing in a boardroom. Sitting beside coordinators, watching their screens, and noting every action they took.
What we mapped:
Morning dispatch assembly. Each coordinator opened four applications: the transport management system (TMS), a shared Google Sheet for driver availability, WhatsApp for real-time driver communication, and Outlook for client delivery confirmations. Information from all four sources was manually combined into a daily dispatch plan.
Route assignment. Routes were assigned based on coordinator experience and driver preference. No systematic optimisation. Senior coordinators knew which drivers performed well on which corridors, but this knowledge wasn't documented anywhere.
Client communication. Delivery confirmations were sent to clients via email, one at a time. Each confirmation required the coordinator to pull the delivery time estimate from the TMS, format it into an email template, and send it. For 60-80 deliveries per day, this consumed roughly 90 minutes.
Exception handling. When a delivery failed, was rescheduled, or encountered a problem, the coordinator updated the TMS, notified the client by email, informed the driver by WhatsApp, and updated the shared spreadsheet. Four separate updates for a single event.
Week 2: Data Analysis and Stakeholder Interviews
With the process map established, we analysed six months of dispatch data and conducted structured interviews with coordinators, drivers, the operations manager, and three key client contacts.
The data revealed patterns the team hadn't noticed. Certain vehicle-route combinations consistently produced better on-time performance. Specific times of day had predictable delay patterns on the N1 and N2 corridors. Client rescheduling requests clustered on Mondays and Fridays.
The interviews revealed frustrations. Coordinators felt they spent too much time on administrative updates and too little on problem-solving. Drivers complained about receiving conflicting information when plans changed. The operations manager lacked real-time visibility into fleet status.
| Metric | Before | After | Change |
|---|---|---|---|
| Processing Time | Manual (hours) | Automated (minutes) | -90% |
| Errors | 8-12% | <1% | -95% |
| Staff Hours/Week | 40+ | 8-10 | -75% |
| Customer Satisfaction | 3.2/5 | 4.6/5 | +44% |
What Did the Audit Uncover?
The investigation identified 9 distinct manual touchpoints in the dispatch workflow and 3 approval steps that added time without adding value.
The 9 Manual Touchpoints
-
Driver availability check -- coordinators called or WhatsApped each driver every morning to confirm availability, despite the TMS having a scheduling module that nobody used because it hadn't been configured for the firm's shift patterns.
-
Route assembly -- manual combination of delivery orders from the TMS with vehicle capacity and driver assignment in a spreadsheet.
-
Route approval -- the operations manager reviewed and approved every daily route plan before dispatch, a process that took 30-45 minutes each morning.
-
Client confirmation emails -- individual emails sent to each client with estimated delivery windows.
-
Driver briefing -- coordinators called each driver individually to communicate their route, rather than using the TMS mobile app (which existed but wasn't adopted due to poor initial training).
-
Mid-day status updates -- coordinators called drivers at midday to check progress, then updated the spreadsheet.
-
Exception logging -- when deliveries failed or were rescheduled, coordinators updated four separate systems manually.
-
End-of-day reconciliation -- coordinators compared the planned dispatch with actual deliveries, manually noting discrepancies in the spreadsheet.
-
Weekly reporting -- the operations manager spent Friday afternoons compiling weekly performance data from the spreadsheet into a PowerPoint presentation for management.
The 3 Redundant Approvals
Approval 1: Daily route approval. The operations manager reviewed every route plan. In practice, he approved 96% of plans without changes. The 4% he adjusted were almost always due to information he had (a vehicle in maintenance, a driver on leave) that should have been in the system already.
Approval 2: Exception re-routing approval. When a delivery needed to be rerouted due to a failed attempt or client request, the coordinator needed operations manager sign-off before making the change. Average wait time for approval: 23 minutes. In an industry where delivery windows are measured in hours, this delay had real consequences.
Approval 3: Overtime authorisation. When late deliveries required a driver to exceed standard hours, the coordinator needed approval from both the operations manager and the finance team. This two-step approval averaged 40 minutes and often resulted in the decision being moot because the delivery window had closed.
What Was the Solution?
The solution wasn't a single AI tool. It was a phased restructuring of the dispatch workflow, with AI capabilities introduced where they addressed specific, documented inefficiencies.
Phase 1: System Configuration and Process Cleanup (Weeks 1-3)
Before introducing any AI, we configured the existing TMS properly. The driver scheduling module was set up with correct shift patterns. The mobile app was reinstated with proper training for all 38 drivers. The shared spreadsheet was eliminated by ensuring all data lived in the TMS.
This alone removed touchpoints 1, 5, and 6. No AI required. Just proper use of existing tools.
Phase 2: AI-Assisted Route Optimisation (Weeks 4-6)
With clean data flowing through a single system, we introduced AI-assisted route optimisation. The system analysed historical delivery data, traffic patterns on the Western Cape and Gauteng corridors, vehicle capacity, and driver performance to generate optimised daily route plans.
This addressed touchpoint 2 (manual route assembly) and, critically, eliminated Approval 1 (daily route approval). The AI-generated plans were consistently better than manual plans. The operations manager shifted from approving every plan to reviewing exception reports, a role that required 10 minutes daily instead of 45.
Phase 3: Automated Communication and Exception Handling (Weeks 7-10)
We implemented automated client notifications triggered by TMS events. When a delivery was dispatched, the client received an automated message with a tracking link and estimated delivery window. When a delivery was delayed or rescheduled, the client was notified automatically.
This eliminated touchpoints 4 (client confirmation emails) and most of touchpoint 7 (exception logging). The system also removed Approval 2 by establishing rules-based re-routing: if the change fell within defined parameters (same driver, same day, no overtime), it happened automatically. Only changes exceeding those parameters required human approval.
Phase 4: Reporting and Oversight Automation (Weeks 11-12)
The final phase replaced manual reporting (touchpoints 8 and 9) with automated dashboards. Real-time fleet status, daily reconciliation, and weekly performance reports were generated automatically from TMS data.
Approval 3 (overtime authorisation) was streamlined to a single-step process with the operations manager only, removing the finance team from the real-time loop while maintaining financial oversight through automated cost reporting.
What Were the Results?
After 12 weeks of phased implementation, the dispatch team's manual coordination time dropped from approximately 132 hours per week (22 hours per coordinator across 6 coordinators) to roughly 78 hours per week.
The key metrics:
- 22 hours per week of manual coordination eliminated across the team
- Dispatch error rate dropped from 12% to under 3%
- Average exception resolution time decreased from 47 minutes to 11 minutes
- Client satisfaction scores improved from 3.6 to 4.3 out of 5 within two months
- Operations manager reclaimed roughly 6 hours per week previously spent on approvals and reporting
The coordinators weren't replaced. They were freed to focus on strategic work: negotiating delivery windows with key clients, analysing performance trends, and managing driver development. The team's role shifted from administrative coordination to operational optimisation.
What Lessons Apply to Other Logistics Companies?
Several principles from this engagement apply broadly across the South African logistics sector.
Investigate Before You Implement
The most impactful improvements, removing the shared spreadsheet, configuring the existing TMS properly, training drivers on the mobile app, required zero AI. If we'd started by recommending an AI platform, we would have built sophisticated technology on top of broken processes. The investigation phase prevented that.
Redundant Approvals Are Hidden Costs
Approval workflows often accumulate over time as responses to past mistakes. Each one made sense when it was introduced. But collectively, they create bottlenecks that nobody questions because "that's how we've always done it." Auditing approval chains specifically is one of the highest-ROI activities in any process investigation.
Existing Tools Are Underused
This firm was paying for a TMS with capabilities they weren't using. Before buying new technology, audit what you already have. In our experience working with South African logistics companies, existing systems typically have 30-40% more capability than teams realise.
Phase the Implementation
Trying to change everything at once overwhelms teams and creates resistance. The phased approach allowed the dispatch team to absorb each change, build confidence, and see results before the next phase began. Each phase's success built momentum for the next.
For a practical guide on transitioning from manual to automated processes, see our manual-to-automated transition guide.
Could Your Operations Benefit From an SOP Audit?
Every logistics company has processes that evolved organically over years. Manual touchpoints accumulate. Approval chains grow. Workarounds become standard practice. The inefficiencies are invisible to people inside the operation because they've always been there.
An investigation-first SOP audit makes those inefficiencies visible, measurable, and addressable. It's the starting point for genuine operational improvement, whether that improvement involves AI or simply better use of what you already have.
Learn more about our approach to process assessment and AI implementation and how an SOP audit could uncover hidden efficiency in your operation. The conversation starts with your processes, not with a product.
Related Resources:
- AI for Logistics and Route Optimisation
- Manual to Automated Transition Guide
- Calculate ROI for Process Automation
Our team at Smart AI Solutions, led by Loxly Atkinson, has implemented these solutions across multiple South African industries.
Frequently Asked Questions
What did the investigation process involve?
The investigation included two weeks of process shadowing (sitting with dispatch coordinators and watching their workflows), six months of dispatch data analysis, and structured interviews with coordinators, drivers, the operations manager, and key clients.
What did the audit uncover?
The audit identified 9 manual touchpoints in the dispatch workflow and 3 redundant approval steps. Most impactful findings: the TMS had unused capabilities, a shared spreadsheet duplicated system data, and approval wait times averaged 23-40 minutes per exception.
What were the results?
After 12 weeks of phased implementation, manual coordination dropped by 22 hours per week. The dispatch error rate fell from 12% to under 3%, exception resolution time decreased from 47 to 11 minutes, and client satisfaction improved from 3.6 to 4.3 out of 5.
What lessons apply to other logistics companies?
Investigate before implementing, audit redundant approval chains specifically, check whether existing tools have unused capabilities, and phase changes rather than doing everything at once. In our experience, existing systems typically have 30-40% more capability than teams realise.
