What is the difference between route planning and route optimisation?

Route planning decides which stops go on which route. Route optimisation decides the order they are driven in. Operations with one vehicle only need optimisation; anyone running several drivers needs both. Geo2 handles allocation and sequencing together.

How much does route optimisation software cost?

Pricing models differ more than headline prices. Some products charge per stop or add overage above a monthly threshold, so volume raises the bill. Others charge a fixed amount per user. Geo2 charges per user per month with no per-stop fees, starting free for one driver and 15 stops per route.

Is there free route optimisation software?

Geo2's free plan includes route optimisation for up to 15 stops per route with camera scanning, voice input and proof of delivery, and requires no card. Most free tiers in this market are time-limited trials, so check whether the free plan expires.

Can route optimisation software handle delivery time windows?

Some can, usually on higher tiers. Geo2 supports required-from and required-to windows per stop on Advanced and Enterprise, and where a window cannot be met it flags this before dispatch rather than refusing to build a route.

Does route optimisation software replace sat-nav?

Not usually, and it shouldn't. Optimisation decides the order of stops; navigation handles turn-by-turn driving. Geo2 has no navigation engine and hands each stop to Google Maps, Apple Maps or Waze.

What is route optimisation software?

Route optimisation software decides the order and timing of a set of delivery stops so a vehicle covers them in the least time or distance. It works across real road networks, not straight lines, and accounts for delivery time windows, vehicle weight and height limits, driver shift lengths and vehicle capacity. For one driver it produces a sequence. For a fleet it also decides which driver gets which stops.

Route optimisation is also spelled route optimization.

In short:

What it does: turns a list of addresses into a driving order, and splits work across drivers
What it uses: real road geometry, traffic data, time windows, vehicle dimensions, shift lengths
What it doesn't do: turn-by-turn navigation, fixing bad address data, predicting traffic with certainty
Typical pricing: per user per month, or per stop — the difference matters more than the headline price
Free options: exist, but most are time-limited trials rather than permanent free plans

What route optimisation software can't do

Fix bad address data. A postcode that geocodes to a village centre rather than the farm produces a confidently wrong route. Address quality caps route quality.
Predict traffic with certainty. Live and historical traffic improve a plan. They don't make it a fact.
Encode local knowledge. The estate where the barrier sticks, the street where nobody answers before ten. Good software lets drivers override rather than pretending the constraint doesn't exist.
Replace navigation. Optimisation decides order; turn-by-turn is a separate job. Products that insist on both usually do one worse.
Rescue the wrong inputs. Wrong stops, wrong vehicle, addresses arriving as a photo of a printout — sequencing isn't the constraint there.

Where Geo2 fits

Geo2 is route optimisation software for delivery drivers and small fleets. It sequences stops on real roads with vehicle weight, height, width and length applied, and hands turn-by-turn driving to Google Maps, Apple Maps or Waze rather than replacing them. Addresses go in by camera scan, manifest photo, bulk paste of up to 50, voice or file import. Pricing is fixed per user with no per-stop charges at any volume.

The free plan covers 15 stops per route with no card and no expiry — enough to test a real morning. Time-window scheduling and multi-driver planning are on Advanced and Enterprise.

Geo2 was built by a team that spent twenty years in supply chain consultancy first, which shows mainly in the unglamorous parts: vehicle constraints, proof of delivery detail, and the decision not to build a navigation engine.

Plan a route free → · Route optimisation for teams →

Why sequencing by hand doesn't work

The number of possible orders grows faster than intuition suggests.

Stops	Possible Orders
5	120
10	3,628,800
20	2.4 × 10¹⁸
50	3.0 × 10⁶⁴

At 50 stops there are more possible routes than atoms in the Milky Way. This is a known problem in operations research called the vehicle routing problem, first described in 1959, and it is why sequencing is automated rather than reasoned through.

In practice most rounds are ordered one of three ways: a depot system hands over a sequence built around the depot's priorities, a planner sorts by postcode, or the driver decides. All three work. All three leave time unclaimed.

The five constraints that separate products

Constraint	What it means	Why it matters
Real road geometry	Distance measured on roads, not as the crow flies	Two stops 400m apart can be a 12-minute drive with a railway between them
Time windows	Required-from and required-to per stop	Without it, promised arrival times are ignored entirely
Vehicle limits	Weight, height, width, length	A 3.5t Luton can't take routes a car can
Driver time	Shift length, start and end points, breaks	Breaks with a real address mean fuel and food are planned, not improvised
Multi-vehicle allocation	Splitting stops across drivers	A different and harder problem than sequencing one route

The last row is the one buyers get wrong most often. Sequencing a single route and allocating work across a team are different problems, and plenty of single-driver tools are sold as fleet software.

On time windows, ask whether they are hard or soft. Hard windows mean the software refuses to build a route if a window can't be met. Soft windows mean it builds the best available route and flags the promise at risk. Soft is more useful at 6am — a plan with one late arrival beats no plan.

Route planning vs route optimisation

The terms are used interchangeably but describe different steps.

	Route planning	Route optimisation
Question answered	Which stops go on which route?	What order should they be driven in?
Input	A day's orders, a list of drivers	One route's stops
Output	Allocated routes	A sequence
Needed by	Anyone with more than one driver	Everyone

Most products do both. If you run one vehicle, you only need the second.

How to choose route optimisation software

Five questions separate products more reliably than a feature list.

1. Does the price move with your volume?

Model	How it works	Effect
Per stop	Charged per delivery, or banded with overage	Your busiest month is your most expensive
Per user	Fixed monthly amount per driver seat	Volume changes nothing

Work out what your best month costs under each, not your average.

2. How do addresses get in?

This decides whether the software gets used at all. Manual entry is the most common reason route planners are abandoned. Look for bulk paste, file import, camera scanning of labels, manifest reading, and an API if you have a system to connect.

3. Does it handle your vehicles and your promises?

Vehicle dimensions if you run anything bigger than a car. Time windows if you promise arrival times. Both are often on higher tiers than buyers expect — check the plan, not the feature list.

4. What happens when the day changes?

Plans survive until about nine in the morning. Can a driver add a stop, skip one, or re-sequence from their current position without a planner rebuilding the route, and without pulling over?

5. Will drivers actually use it?

The planner's opinion isn't the one that matters. Software that needs a signal, runs slowly, or replaces a navigation app drivers already trust produces routes on paper that aren't the routes driven.

How AI changed route optimisation

Three things genuinely changed. One didn't.

Changed	What it enabled
Machine reading	Labels, manifests and screenshots read by camera, removing manual entry
Voice control	Changing a route mid-shift without stopping
Duration prediction	Stop times learned from history — a tower block isn't a house with a driveway

Unchanged: the sequencing itself. The algorithms solving the vehicle routing problem are decades-old operations research, refined rather than replaced. A product describing its routing engine as AI is usually describing a rebrand. The honest claim is that AI removed the barriers around optimisation, not that it replaced the maths.

What is Route Optimisation Software?