Frequency is the only transit statistic that matters to a passenger

This works through service frequency in the order the parts actually depend on each other.

Frequency is freedom

When a service runs every few minutes, passengers stop planning around it and simply turn up, which removes an entire category of friction. The commonly cited threshold is around ten minutes, below which timetable consultation largely stops. That change in behaviour is much larger than the arithmetic saving in waiting time suggests.

It is also what makes transfers viable, because a missed connection costs minutes rather than a half hour.

Waiting feels worse than moving

Transport research consistently finds passengers weight waiting time at roughly double in-vehicle time. This means a slower service that comes twice as often can be preferred to a faster infrequent one, even when total journey time is identical. Investment appraisals that value all minutes equally systematically misprice frequency improvements.

A large part of the penalty is uncertainty rather than duration, which is why a countdown display reduces perceived waiting although the vehicle arrives at precisely the same moment, and why shelter, seating and lighting move the number as much as the timetable does.

Reliability compounds with frequency

On a frequent route a single late vehicle is absorbed; on a half-hourly route it is a thirty-minute failure. Bus bunching — where vehicles catch each other and arrive in pairs — is a frequency-management problem that destroys the passenger experience.

Holding strategies, signal priority and dedicated lanes exist mainly to protect headway regularity rather than raw speed. The causes are almost always outside the vehicle — congestion at a handful of junctions, slow boarding where fares are paid in cash to a driver, and a schedule written with no recovery time at the terminus — which is why padding a timetable treats the symptom and priority measures treat the cause.

The network shape follows from it

High frequency is expensive per route, which forces a choice between a dense infrequent network and a sparse frequent one. Cities that chose the frequent grid with easy transfers generally show higher ridership per vehicle hour than those that chose coverage. That is an explicit political trade-off between ridership and geographic equity, and it is worth naming as one rather than treating as a technical decision.

Across a network, the usual compromise is a frequent core with infrequent feeders, and it only works if the transfer is designed properly, because a passenger who has to wait twenty minutes at an interchange has been handed the sparse network with extra steps in it.

What passengers actually notice

A published frequent-service map, showing only routes above a stated frequency, changes perceived usability immediately. Several cities have done this and reported ridership increases with no service change at all. Legibility is part of the service, and it is far cheaper than infrastructure.

Across a network, it fails in specific and fixable places: a stop with no route number on the pole, a service that changes its number partway along, a fare rule that charges twice for one journey with a transfer in it, and a schematic map that hides which streets a bus actually runs down.

The data here is patchy — most cities do not publish it consistently.

Span decides who the service is for

A route running every five minutes at rush hour and stopping at seven in the evening is unusable for anyone whose shift ends at ten, which describes a large share of the workforce least able to drive. A trip is a chain, so the return leg governs the whole decision: people will not take a bus in the morning if they cannot rely on getting home at night. Evening and weekend service is the first thing cut when budgets tighten, because those hours carry the fewest passengers and the saving therefore looks best per trip removed.

The cost of those hours is not proportional either, since the vehicles and depots are already paid for and the marginal expense of an evening hour is mostly the driver.

The takeaway

Ask how often it runs. Everything else about a service is downstream of that.

The design decision is visible long after the people who made it have gone.