How cities can introduce shared electric scooters

Cities that had a bad experience with shared scooters usually share one detail: the fleet arrived before the rules did. Cities that had a good experience usually reversed that order. This is what the second sequence looks like.

Decide what problem you are solving

Scooters can serve very different purposes: connecting districts, extending the reach of public transport, moving visitors between attractions, reducing short car trips in a congested centre. These lead to different deployment areas and different fleet sizes.

A city that has not decided which of these it wants will find it hard to judge whether the service is working, because it has no measure of working.

Draw the map first

Before any vehicle is deployed, four things should exist on a map and be agreed:

1. 01The riding area: the boundary outside which the service does not operate.
2. 02No-ride areas: pedestrian streets, private grounds, hazardous roads.
3. 03Parking areas: where a ride may legally be ended.
4. 04Speed-limited areas: promenades, squares, school approaches, market streets.

Agreeing this afterwards is far harder, because by then riders have habits and the operator has a fleet distributed around the old assumptions.

Ask the operator the enforcement question

This is the question that separates operators, and it is worth asking plainly: when a rider enters a no-ride zone, what happens?

If the answer is that the rider receives a notification, the city has been offered a rule that depends on rider compliance. If the answer is that the vehicle stops delivering power, the city has been offered a rule.

Ask the same question about parking. Can a ride be ended in a no-parking area, or is it refused? “We fine the rider afterwards” is a different and much weaker answer than “the ride does not end”.

Start small on purpose

A limited first deployment in a defined area is not timidity; it is how you find the problems at a size where they can be fixed. You will learn where vehicles accumulate, which streets generate complaints and where parking areas are in the wrong place, and all of that is cheaper to learn with a small fleet.

Set the terms you will need later

Things worth fixing in writing at the start, because they are difficult to add afterwards:

• Who is responsible for retrieving an incorrectly parked vehicle, and how quickly.
• What data the operator provides to the city, and in what form.
• How the zone map is changed, and how quickly a change reaches the vehicles.
• Rider identification requirements.
• What happens to the vehicles if the operator ceases trading.
• Insurance and liability.

Measure the right things

Ride counts flatter everyone and tell you little. More useful:

• Where trips start and end, which is planning data you will not otherwise have.
• What proportion of rides end in a permitted parking area first time.
• Complaints per thousand rides, and what they are about.
• How many vehicles are unavailable at any time, and why.
• Whether scooter trips are replacing car trips, which needs asking riders, not reading logs.

Expect to change the map

The first zone map will be wrong somewhere. A street will turn out to be busier than expected; a parking area will be inconveniently placed and quietly ignored. This is normal, and the important thing is that changing the map is quick and does not require touching the vehicles.

A city that can redraw a boundary in the morning and have it in force that afternoon is in a fundamentally different position from one that has to negotiate a fleet update.