Delivery Route Optimization Service

What a delivery route optimization service does

A navigation app solves a simple problem: how to get from point A to point B. A delivery route optimization service solves a different and much harder one: how to split dozens or hundreds of addresses among several drivers or couriers, and in what order each of them should visit their stops, so that every order is delivered on time at the lowest cost.

In theory this is known as the vehicle routing problem (VRP). What makes it tricky is that the number of possible combinations explodes: even with a few couriers and a couple of dozen addresses, you can't try every option by hand or eyeball it. That's why an optimizer relies on specialized algorithms that find a near-optimal solution in a reasonable amount of time.

The output isn't a single line on a map but a ready-made plan for the day:

• Order assignment — which addresses go to which courier or vehicle.
• Stop sequence — the order of stops within each route.
• Estimated arrival time at every stop, including the time needed to complete the order.
• Unassigned orders — those that don't fit the day's resources, with a clear reason why.

That last point is often underrated. A good optimizer doesn't hide the problem; it tells you plainly that this order doesn't fit its time window or the vehicle capacity, so you can decide what to do with it.

Which constraints route planning software must handle

A route that looks short on the map is useless if the courier arrives at a locked door or can't fit all the goods in the vehicle. So the value of route planning software comes down to which real-world constraints it can take into account.

• Delivery time windows. The customer expects the order between 10 and 12, the office only accepts deliveries until 6 pm, the warehouse starts dispatching at 9 am. A missed window means an unhappy customer or a second trip.
• Working hours. Shifts start and end at set times, and they can differ from person to person.
• Capacity. The weight, volume or number of parcels that fits in a van, a courier bag or a car trunk.
• Service time. Handing over a parcel takes minutes; an installation or a site survey can take an hour or more. Without it, estimated arrival times quickly drift away from reality.
• Start and end points. Leaving from the warehouse, from the worker's home or from another location; returning to base or finishing at the last stop.
• Priorities. Urgent and important orders should be scheduled first when there aren't enough resources for everything.

If a system only looks at distance, the dispatcher ends up manually fixing every route, and the benefit of automation disappears. When comparing options, check against the constraints of your own business rather than a generic promise to "optimize routes". For more on working with time windows, see our article on delivery time windows.

How multi-stop route calculation works

In any serious system, calculating a route across multiple stops goes through several stages, and each one affects the quality of the result.

• Address preparation. Addresses are converted into coordinates. A mistake at this step — the wrong building or even the wrong city — breaks the whole plan, so it helps to see the stops on a map before running the calculation and fix anything that looks off.
• Distance and time matrix. For every pair of stops, the system estimates the travel time between them. Straight lines on a map won't do here: what matters is the actual road network, one-way streets and bridges.
• Optimization. The algorithm assigns orders to workers and chooses the visiting order so that all constraints are met and total mileage or time is minimized.
• Review and adjustment. The dispatcher reviews the result and, if needed, pins an order to a specific courier or moves it manually and recalculates.

Changes during the day are a separate question. A new urgent order, a cancellation, a courier calling in sick, a traffic jam — a rigid morning plan is out of date by lunchtime. So check how easy it is to re-optimize routes when some orders are already completed and your workers are scattered across the city.

Online route optimizer vs. delivery management system

There are two classes of solutions on the market, and it's worth telling them apart before you start choosing.

An online route optimizer is a standalone tool: upload a list of addresses, get routes, send them to your couriers. This works for one-off jobs or small volumes. The catch is that everything that happens to an order after the route is built stays outside the system: statuses, confirmations, photo reports and customer communication still run through messengers and phone calls.

A delivery management system builds optimization into the full workflow:

• importing orders from your ERP, CRM or a spreadsheet;
• assigning orders to workers and building routes;
• sending the route to the courier's mobile app;
• tracking progress on a live map;
• confirming delivery on site with photos and a checklist;
• showing order statuses to the customer in a client portal.

If delivery is a daily process rather than a one-off task, the second option almost always pays off: a route that nobody monitors once the courier leaves quickly stops matching the plan, and the optimization loses its point.

How to choose a service and test it on your own data

Route optimizer pitches all look alike: pretty lines on a map and promises of savings. To compare solutions on substance, set up your own test.

• Real data. Take orders from a typical day and from your busiest day — with actual addresses, time windows and volumes.
• Baseline. Record how those days were planned by hand: how many workers went out, total mileage, how many windows were missed, how long planning took.
• Run it through the system. Ask the vendor to plan the same days with your constraints and compare the result against the baseline using the same metrics.
• Change scenarios. Add an urgent order mid-day, remove a courier, shift a time window — and see how quickly the system rebuilds the plan.
• The worker's experience. Check what the route looks like on the courier's phone and how many taps it takes to close an order.
• Integrations and rollout. Find out how orders get into the system, how long implementation takes and whether there are long-term commitments.

A test like this takes a day or two and tells you more than weeks of emails — and gives you hard numbers to base the decision on: your own, not the marketing ones.

Route optimization in itlogist

itlogist is a system for managing couriers and field workers where route optimization is built into the daily workflow rather than being a separate tool. The optimization engine is built on OR-Tools and takes delivery windows and constraints into account, and finished routes appear on a live map right away.

Orders are received and assigned to workers in a single screen, with data coming from 1C, AmoCRM, Bitrix24 or Excel. Couriers get their route in a mobile app with photo reports and checklists, or in a mobile web interface with no app install required. Delivery is confirmed on site, and the customer sees order statuses in their client portal. Learn more on the itlogist route planning page.

The system is designed for teams of 5 to 100 workers, rollout takes 7 days, and there are no long-term contracts. A sensible way to start is a pilot at one warehouse or with one group of couriers: compare the results with manual planning on your own data, and only then move your entire volume over.