# Route optimization: in what order should you visit your customers?

> Route optimization explained: why stop order is such a hard puzzle, which constraints count for field crews, and a mileage calculation for a small business.

Updated on October 15, 2026 · https://app.pilotiko.com/en/glossary/route-optimization

Route optimization is planning each vehicle's itinerary to set the order of customer visits or pickups at the lowest cost in distance or time. It has to respect real-world constraints: arrival windows promised to customers, vehicle capacity, how long each job takes and the hours each crew can work.

## In short
- The OQLF's recommended French term is « optimisation de tournées de véhicules ».
- With 10 stops there are already 362,880 possible visit orders.
- Arrival windows promised to customers change the whole calculation.
- In Pilotiko, route optimization is planned and not offered yet.

## What do the terminology references call it?

Quebec's Grand dictionnaire terminologique pairs the French « optimisation de tournées de véhicules » with the English vehicle routing and scheduling. It defines the concept as planning vehicle itineraries so as to set the sequence in which each customer is visited or each pickup point is collected, and notes that it applies to road vehicles as well as automated guided vehicles [1]. In everyday English, people say route optimization or route planning.

The key word is **sequence**. Choosing the technician is [dispatching](https://app.pilotiko.com/en/glossary/dispatch); choosing the order in which that technician drives from stop to stop is route optimization.

## Why is stop order so hard to compute?

The textbook case is the travelling salesperson problem: find the shortest trip that visits every customer once and returns to the start. Google's OR-Tools documentation shows the scale involved: with 10 locations (not counting the starting point) there are 362,880 possible routes, and with 20 locations the count jumps to 2,432,902,008,176,640,000. Trying every option only works for tiny sets, so routing software uses search techniques that find an optimal or near-optimal answer [2].

The same page lists, among typical examples, a cable TV company assigning routes to technicians for residential service calls [2]. That is precisely the day of a plumber, a pest-control technician or an appliance repair person.

## Which constraints matter for a service business?

With several vehicles, the task becomes the vehicle routing problem, usually with constraints such as a maximum load per vehicle or visits that must fall inside time windows requested by customers [2]. That last variant is called the vehicle routing problem with time windows, and it is solved from a matrix of travel times rather than travel distances [3].

| Constraint | Real-life example | Effect on the calculation |
|---|---|---|
| Promised arrival window | Customer home only from 8 to 10 a.m. | Fixes the order of some stops |
| Job duration | Full treatment: 90 min; follow-up: 20 min | Shifts arrival times for every later stop |
| Vehicle capacity | Number of water heaters the truck can carry | May force a trip back to the shop |
| Skills | Only one tech holds the required certification | Removes possible pairings |
| End of shift | Back at the yard by 5:30 p.m. | Caps the number of stops per day |
| Emergency | Flood reported at 11 a.m. | Forces a mid-day recalculation |

The more constraints you add, the closer the plan gets to reality, and the more you depend on accurate addresses, durations and windows.

## What makes a route “optimal”?

It depends on the goal you set. Minimizing total distance sounds natural, but with no other constraint the best answer is to give every stop to a single vehicle. OR-Tools suggests minimizing the longest single route across all vehicles instead, which fits when you want all visits finished as early as possible [4].

A small business may care about something else: keeping every promised [time slot](https://app.pilotiko.com/en/glossary/time-slot), limiting overtime or keeping one tech free in the afternoon for emergencies. Settle the objective before you compare tools.

## How many kilometres can stop order change?

**Calculation: one day for a pest-control technician in Quebec City**

- Stops booked for the day : 8
- Distance following the order calls came in (assumption) : 142 km
- Distance after grouping by neighbourhood, same list (assumption) : 96 km
- Kilometres avoided (142 − 96) : 46 km
- Driving time at a 45 km/h average: 142 ÷ 45, then 96 ÷ 45 : 3 h 09 vs. 2 h 08
- Driving time freed up : about 1 h a day
- **= In this scenario, reordering the same 8 stops frees up roughly an hour of driving per day without adding a single customer.**

## Which route optimization mistakes should you avoid?

- Optimizing kilometres while ignoring promised windows, so the customer waits and the visit fails.
- Using fixed job durations when some tasks take anywhere from 20 minutes to an hour.
- Feeding in incomplete or badly geocoded addresses that throw off the whole plan.
- Reshuffling the route without telling the tech or the customer, which feeds [no-shows](https://app.pilotiko.com/en/glossary/no-show-rate).
- Filling every minute, with no buffer for traffic or an emergency call.

## Does Pilotiko optimize routes?

What is available: [scheduling](https://app.pilotiko.com/en/product/booking-and-scheduling) already proposes who should take each booking based on calendars and workload, using real open time slots, with a person confirming the choice. Territory and stop order remain a human decision for now.

> Coming soon. Route optimization is not built in Pilotiko yet: no itinerary or stop-order calculation is offered today.

## Frequently asked questions

### What is the difference between route optimization and dispatching?

Dispatching decides who takes which call. Route optimization decides the order and path of each vehicle's visits. They work together: a perfectly ordered route is useless if the wrong tech is sent.

### How many daily stops before optimization is worth it?

With three or four stops in a small area, an experienced person often finds a good order by eye. The number of combinations grows very quickly, though: 10 stops already give 362,880 possible orders. The more vehicles, stops and time windows you juggle, the more a calculation helps.

### Isn't a navigation app enough?

A navigation app finds the best path between two points, or sometimes through a short list of stops. It generally ignores promised windows, job durations and how work is split across several vehicles. For a fleet, you need a routing tool.

### Should a route be recalculated during the day?

Yes, when an emergency, a cancellation or a major delay comes up. Let every affected customer know their new arrival time. A silent reshuffle creates more problems than it solves.

### What data does route optimization need?

Accurate addresses, realistic durations for each type of job, promised arrival windows, crew start and end times and vehicle capacity. Without that data, even the best algorithm will produce a route nobody can drive.

## Sources

1. Office québécois de la langue française, Grand dictionnaire terminologique, « optimisation de tournées de véhicules », https://vitrinelinguistique.oqlf.gouv.qc.ca/fiche-gdt/fiche/2068273/optimisation-de-tournees-de-vehicules (accessed on 2026-10-15)
2. Google for Developers, OR-Tools, « Vehicle Routing », https://developers.google.com/optimization/routing (accessed on 2026-10-15)
3. Google for Developers, OR-Tools, « Vehicle Routing Problem with Time Windows », https://developers.google.com/optimization/routing/vrptw (accessed on 2026-10-15)
4. Google for Developers, OR-Tools, « Vehicle Routing Problem », https://developers.google.com/optimization/routing/vrp (accessed on 2026-10-15)
