Route Optimizer
Deliveries, viewings, errands — the order you visit stops decides the day. Add your stops and the optimizer trials the visiting order to cut the total road distance, then routes the improved sequence.
By Aryan Khan, Founder & Editor · Updated October 2026
For example, five New York stops visited in the order they were typed can run about a quarter longer than the same stops in geographic order — the optimizer's whole reason to exist, illustrated below.
How to use Route Optimizer
The order of stops is a cost decision hiding in a list. Let trials decide it:
The Route Optimizer takes your start and every stop, trials visiting orders to cut the total road distance, and routes the improved sequence with its legs and total. Add the stops as they occur to you — the optimizer's trials exist because typed order is rarely the road's order.
- 1Add your start and every stop.
- 2Switch optimisation on (it is, on this page).
- 3Read the reordered route and its distance.
What people use Route Optimizer for
1. Delivery rounds
A quarter off a daily round is fuel, hours and a second round's capacity. The Multi-Stop Route Planner routes a fixed order; this page improves the order itself.
2. Service and viewing days
Engineers, agents and inspectors chain addresses across a county. The optimised sequence is the difference between a day that fits and one that frays.
3. Errand Saturdays
Six stops typed in worry order become four fewer backtracks in road order. The saving is Saturday itself.
4. Checking your instinct
Think your order is optimal? Run it against the trials. Being right is satisfying; being wrong was costing you miles.
Worked examples
Illustration with straight-line distances between five New York stops (Manhattan, the Bronx, Queens, Brooklyn, Staten Island, Jersey City): typed order ≈ 116 km of travel; geographic order ≈ 87 km — about 25% less, before roads make both longer but keep the lesson.
Start fixed at the depot, stops trialled freely, return priced in: the optimizer's sequence is a rota a driver can simply follow — legs and total included.
What optimisation changes
The same stops, before and after — reading guide for the optimizer's output, using the illustration's proportions.
| Figure | Typed order | Optimised order |
|---|---|---|
| Stops changed | — | None — same list, better sequence |
| Total distance | The long way round | Cut, often by a fifth or more |
| Leg pattern | Backtracks and crossings | A sweep across the map |
| Driver brief | A list | A sequence with legs and times |
| Start point | Fixed | Fixed — trials move the middle, not the anchor |
Exact optima for many stops are famously hard in general; the optimizer trials orders heuristically and returns a strong improvement, honestly labelled as such by its totals.
The best order for your stops, solved
Humans are surprisingly bad at the travelling-salesman problem, and surprisingly confident about it. Type your stops in any 'logical' order and this tool shows what the routing engine knows: reordering the intermediate stops on the real road network typically shaves 15–30% off the total distance, because intuition underweights cross-town backtracking that the engine prices exactly. Your first stop stays fixed as the origin — the depot, the office, home — and the rest fall into the sequence that minimises driving.
The reordered route draws on the map with the new total, so the saving is visible rather than asserted, and a quick visual scan catches the classic failure mode of bad optimisation (a criss-crossing path) before it costs you a morning. For a dozen stops the engine's heuristic is effectively optimal; the page says plainly that fleet-scale problems belong to dedicated logistics software. It is the analytical heart of the delivery planner and the multi-stop route planner, offered bare: bring your stops, leave with the order. Export happens in the sibling tools, and the shareable URL keeps the solution reproducible for whoever drives tomorrow.
How to interpret the result
The most common errors come from using a correct result to answer the wrong geographic question. Keep these points in mind for Route Optimizer:
- Add a congestion buffer to any free-flow travel time: 20–40% in peak urban areas is a sane rule of thumb, and the interface labels its times as free-flow for exactly this reason. For appointments, plan with the buffer; for physics, trust the raw number.
- Place route endpoints on roads, not rooftops. Routing engines snap points to the nearest drivable way, and a point in a river or a courtyard can snap somewhere surprising — nudging the pin to the nearest street makes results stable and explainable.
Limits & appropriate use
Routing's limits live in three places: data vintage, access reality and time modelling. The network is OpenStreetMap's current picture — new interchanges lag, private gates may be missing, seasonal ferries keep their own calendars. Access rules reflect mapped law, not today's roadworks. And durations are free-flow physics: speed limits and road classes without your city's sociology, so peak hours, weather and parking belong in your buffer, not in the engine's promise. The tools label all three limits on every result, because an unlabelled estimate is a trap.
What the stack does brilliantly is the reproducible core: the same stops, mode and engine give the same route to everyone, everywhere, which makes it a superb baseline for comparison, screening and planning. Escalation is domain-specific and well understood — professional dispatch adds live traffic and driver hours; logistics tenders add contracted networks; navigation products add certified maps. A free browser tool that hands you a clean, exportable baseline with its assumptions printed is not competing with those; it is feeding them.
- Live congestion and ETAs → traffic-aware commercial routing.
- Driver-hours and windows → transport-management systems.
- Certified navigation → licensed map products with update guarantees.
Data & methodology
Greedy plus improvement passes over routed leg distances, then a final Valhalla/OSRM route through the winning order.
For source details, licences and model limitations, see Data Sources and Methodology.
Key terms worth knowing
- Costing model
- The routing engine's mode profile (auto, pedestrian, bicycle) with its speeds and rules.
- Free-flow time
- Travel time from speed limits and road classes, without congestion.
Editorial guidance: MapBench explains the method and limitations so the tool can be used as a transparent reference, not as a black-box answer. For legal boundaries, safety-critical routing, surveying, engineering, regulated decisions or official statistics, verify the result with the responsible authority or a qualified professional.
Category: Routing & Travel Time. Results are intended for everyday analysis and planning unless a page explicitly states otherwise.
Method, assumptions & limits
- This tool’s method
- Greedy plus improvement passes over routed leg distances, then a final Valhalla/OSRM route through the winning order.
- Before relying on the answer
- Check the stop order, travel mode and whether the result is a network route or a simplified estimate. Repeat the calculation after changing any of these inputs.
- What the result cannot establish
- Road access, congestion and closures can change. An estimated duration is not a live arrival-time commitment.
Coverage: Worldwide. Search, basemap and external data requests may use third-party services; browser-based does not mean every operation is offline. Read the source and licence notes, calculation methodology and privacy policy. If a result looks inconsistent, report a correction with the inputs and mode used.
Frequently asked questions
Clear answers about the calculation, accuracy and appropriate use.
How many stops can I add?
Up to ten, which keeps the order search instant and the day realistic.
Does it fix the start point?
Yes — optimisation keeps your first stop as the start and reorders the rest.
Does the route include live traffic?
Unless the tool explicitly says otherwise, travel times are planning estimates and should not be treated as live-traffic predictions.
Why is road distance longer than straight-line distance?
Roads must follow the available network, terrain, bridges and restrictions, while straight-line distance measures the geometric minimum between two points.
Can route results change over time?
Yes. Road-network data, access restrictions and routing datasets can change, so recheck important journeys close to travel time.
Why is road distance longer than straight-line?
The optimised total is road distance between stops in sequence — the figure your fuel and hours will actually meet.
Is the result guaranteed optimal?
It is a strong heuristic improvement with its distance printed — judge it by the number, not by a promise no honest tool could make for every input.
Timed appointments?
Stops with fixed times constrain the order in ways distance trials do not know. Optimise the flexible stops; pin the appointments in the Multi-Stop Planner.
Traffic-aware ordering?
Not here — order is chosen on distance. If one stop has a hard appointment time, pin it in the planner and optimise the rest.
Does it handle one-way streets?
The routing beneath does — legs are real road legs, one-ways included, so the sequence's total is drivable.
My round has 40 stops.
Split the day into two optimised rounds of 20. Two good sequences beat one impossible list.
Related tools and resources
Fixed orders are routed by the Multi-Stop Route Planner; documents come from the Route Roadbook Generator; and the round's costs from the Delivery Cost per KM Calculator.
More free tools
Sources, methodology & editorial trust
- ✓ Browser-first processing where supported
- ✓ External sources are documented
- ✓ No account required for public tools
- ✓ Review data sources
Calculations are documented with their assumptions, limitations and source context rather than presented as unexplained answers.
Reviewed and maintained by Aryan Khan, Founder & Editor of MapBench.
Part of our routes & travel time hub
Route Optimizer sits in the routes & travel time cluster, alongside 62 other tools that answer neighbouring questions with different methods or datasets.
You may also need
When this answer is only part of the job, these clusters cover the next step.
Read the guide behind this tool
Planning a Road Trip With Nothing but Free Map Tools — A complete weekend-warrior workflow: stops, optimized order, fuel budget, drive-time sanity checks and a GPX file for the car — all without a single paid app.