Route optimisation software is one of the most widely promoted capabilities in fleet management, and one of the most poorly understood. Vendors routinely claim 20-30% cost reductions. Fleet operators who have implemented it know the reality is more nuanced: more useful in some areas than expected, and less transformative in others.
What route optimisation software actually delivers
At its core, route optimisation takes a set of delivery points, a set of vehicles, and a set of constraints (capacity, time windows, driver hours, vehicle type) and calculates the most efficient sequence and assignment. A human planner working with a spreadsheet and a map can do this for a handful of stops. The software does it for hundreds simultaneously, in under 60 seconds, and recalculates when conditions change.
The evidence for what this delivers is consistent across multiple industry sources. Routing software typically reduces distribution costs by 10-20% through lower fuel use, reduced driver time, and fewer vehicles required.[1] Fleet routing technology can result in 5-15% more productive routes and improved customer service.[2] Fleets that fully implement route optimisation report 10-15% reductions in fuel consumption (a pattern consistent with the findings in our fleet fuel management analysis).[3] In Thailand, an express logistics company reported a 60% increase in vehicle utilisation and 30% reduction in operational costs after implementing dynamic routing.[4]
These are real numbers from real deployments, but they represent the upper range. The results your fleet achieves depend on what you are comparing against and how much inefficiency exists in your current planning.
How much is your fleet losing to fuel leakage? Industry benchmarks estimate 10-15% of fleet fuel spend is lost to inefficiency, waste, and unaccounted use. Our calculator shows you where it is going, category by category.
Where route optimisation gains come from
Route optimisation does not create a single dramatic improvement; it accumulates small efficiencies across every route, every day, and the gains sit in four areas.
Sequence optimisation. The order in which a driver visits stops matters more than most operators realise. According to the 2024 Fleet Technology Trends Report, fleets using GPS tracking and route optimisation report a 9% reduction in fuel costs and a 10% reduction in labour costs.[5] Across a fleet of 20 vehicles running daily, those percentage points add up to substantial fuel and time savings.
Workload balancing. Manual planning tends to overload experienced drivers and underutilise newer ones. The software distributes stops based on capacity, location, and time constraints, which means more consistent driver behaviour metrics and fewer overtime costs.
Time window compliance. When customers specify delivery windows, the planning complexity increases exponentially with each additional constraint. A human planner making trade-offs in their head will miss opportunities that the algorithm catches, because the algorithm evaluates all constraints simultaneously rather than sequentially.
Recalculation speed. When a vehicle breaks down, a priority order arrives, or traffic conditions change, the software recalculates the remaining routes in seconds. A dispatcher handling the same disruption manually spends significant time coordinating by phone and redistributing stops, time that directly delays the rest of the day's deliveries.
What the software does not do
This is where most vendor content stops. The limitations matter as much as the capabilities, because ignoring them is how implementations underdeliver.
It does not know your customers. The algorithm sequences stops by location, time window, and capacity. It does not know that the loading bay at Building C does not open until 8am, that the security guard at Warehouse 7 requires a specific form, or that Mrs Tan always needs a phone call 30 minutes before delivery. This knowledge lives in your drivers' heads, and no amount of algorithmic sophistication replaces it.
It does not account for conditions outside the data. In Southeast Asia, this is not a minor gap. Monsoon flooding that closes a side street for three hours, extreme heat that makes a loading dock inaccessible in the afternoon, informal road closures for local events: these conditions affect route viability in ways that traffic data alone cannot capture. Manila's congestion rate sits at 71% and Jakarta's at 53% according to TomTom's Traffic Index.[7] Those are city-wide averages, and the reality on specific streets at specific times is far more variable.
It does not replace experienced dispatchers. Route optimisation shifts the dispatcher's role from route construction to route supervision. The software builds the plan; the dispatcher validates it against operational reality. Operators who bypass the dispatcher entirely, trusting the algorithm without human oversight, often find that failed deliveries increase because the system cannot account for the judgement calls that experienced dispatchers make instinctively.
It does not fix upstream problems. If your address data is inaccurate, the algorithm will route to the wrong locations efficiently. If your vehicle capacity data is outdated, it will overload trucks systematically. If your customer time windows are aspirational rather than contractual, the software will plan around constraints that do not exist. Route optimisation amplifies the quality of your input data. If the inputs are poor, the outputs are poor, just faster.
Route planning is one half. Fleet visibility is the other. SAAN Track provides real-time GPS monitoring, driver behaviour scoring, and route compliance reporting across your entire fleet.
What separates fleets that get value from those that do not
The pattern across successful implementations is consistent: the fleets that benefit most from route optimisation are the ones that treat it as a planning tool within a managed process, not as a replacement for the process itself. Three practices distinguish them.
They feed driver corrections back into the system. When a driver reports that a particular stop is inaccessible at a certain time, or that a road is unreliable during monsoon season, that information is logged as a constraint in the planning system. Over time, the algorithm becomes more accurate because it incorporates local knowledge that no dataset contains. The software gets the fleet efficiency; the drivers get the fleet reliability; the feedback loop gets both.
They measure planned versus actual. Route optimisation produces a plan. Whether drivers follow that plan is a separate question. Fleets that compare planned routes against actual routes driven (using GPS data) discover where the algorithm's assumptions break down and where drivers are deviating for reasons worth investigating. The scale of what is recoverable is significant: UPS's ORION route optimisation system saved 100 million miles annually by 2024 through planned-versus-actual monitoring and continuous recalculation.[6]
They set realistic expectations. Vendor claims of 20-30% cost reduction typically compare optimised routes against the worst-case manual scenario. If your fleet already has competent dispatchers and relatively efficient routes, the gains from software will be more modest, likely 10-15% on fuel and a similar improvement in delivery density. That is still significant at scale, but it is not the transformation that some sales presentations promise.
What the evidence supports
Route optimisation is one of the most justifiable technology investments a fleet operator can make. The evidence consistently supports 10-15% fuel savings, meaningful improvements in delivery density, and dramatic reductions in planning time. For a fleet spending S$50,000 monthly on fuel (to use an illustrative figure; calculate your own fleet's fuel leakage), even a conservative 10% improvement recovers S$5,000 per month, which exceeds the cost of most software implementations within the first year.
What it is not is a standalone solution. The software is only as good as the data it receives, the constraints it is given, and the feedback loop that connects the plan to what actually happens on the road. Operators who understand this see returns that grow over time as the system learns from real conditions. Those who expect the algorithm to solve their logistics on its own are the ones who call it a disappointment 12 months later.
If you are evaluating route optimisation for your fleet SAAN Plan handles multi-constraint route planning for fleets across Southeast Asia and the Middle East. Have a look and see if it fits.
Sources
[1] Paragon Routing, 'A Behind-the-Scenes Look at Manual Delivery Route Planning', 2024.
[2] Descartes, 'What is Route Optimization'.
[3] AJOT, 'How Rising Diesel Prices Are Reshaping Last-Mile Delivery Strategies', November 2025.
[4] FarEye, Thailand Express Logistics Case Study.
[5] Verizon Connect, '2024 Fleet Technology Trends Report'.
