How to Compare Shuttle Programs and Software: Cost, Schedule, Fleet Mix and AI Routing Claims

How do you compare shuttle programs fairly?

Most comparisons fail because each option is described in its own terms: one quotes per month, another per trip, a third per vehicle. The fix is a single scorecard where every option is converted into the same four measures for the same twelve months.

Give each measure a weight out of 100 before you look at any vendor, so the weights reflect your problems and not a sales deck. A hotel group with a flat schedule might weigh cost 40, schedule fit 15, process access 25 and fleet mix 20. A university with changing term timetables might weigh schedule fit highest.

How do you calculate the real 12-month cost?

Add every line a vendor can charge, then multiply by twelve. Do not stop at the headline price. The lines that change the answer are setup, the monthly fee, a per-booking or per-trip fee, any per-vehicle or per-driver fee, and paid support or account management.

Here is the arithmetic using published Shuttle.Software prices, which carry no booking commission. Starter is $99 a month plus a $399 one-time setup, so $99 × 12 + $399 = $1,587. Growth is $199 a month plus $499, so $199 × 12 + $499 = $2,887. Fleet is $349 a month plus $699, so $349 × 12 + $699 = $4,887. Starter covers 1–3 vehicles, Growth 4–12 and Fleet 13–30.

Now the illustrative contrast, using a made-up figure rather than any named vendor. If an option charged a 15% commission and your operation took $40,000 in bookings in a month, that would be $6,000 for that month alone, before any monthly fee. The point is not that every option charges this, but that a per-booking charge grows with your success while a flat plan does not. Put your own volume into the scorecard.

• Ask for the quote in writing with every fee line named.
• Model a quiet month and your busiest month, not an average.
• Count staff time: a free tool that needs a part-time administrator is not free.
• Check the exit cost: can you export bookings, customers and routes as files?

How do you score schedule fit?

Schedule fit is whether the software matches how your demand actually moves. Test three situations with each vendor, using your own data: a normal weekday, your single busiest day and a change made one hour before departure.

For fixed loops, check that published departure times stay fixed while on-demand stops are sequenced around them. For recurring riders, check that a weekly pattern can be set once. For changes, check that a delay or swap updates the driver and the passenger without a phone call. Shuttle.Software works backwards from the drop-off deadline to set pickup times and supports recurring passes; see route optimization and recurring passes for what each does.

• Weekday: does it load your usual timetable without manual re-entry?
• Peak day: does a full run stop selling instead of overbooking?
• Late change: how many steps to move a pickup and notify everyone?
• Seasonal change: can you switch a timetable on by date?

What is process access, and why does it decide adoption?

Process access is how easily the right person can do the right task: a rider books, a manager approves, a dispatcher reassigns, a finance clerk exports. Programs often fail here, not on price, because the tool is hard for non-specialists.

Ask for the role list and watch each role complete its main task. Count steps and time them. For a corporate or campus program, check that an account manager can see only their own cost centre, and that a month-end report can be produced without a support ticket. Shuttle.Software provides corporate and hotel accounts and monthly account invoicing on the Growth and Fleet plans.

Can one system reserve buses, vans and cars together?

Yes, if the platform treats the vehicle class as a setting with its own seat count, rules and price, not as one generic car. A mixed fleet breaks when a rule written for an eight-seat van is applied to a fifty-seat coach.

Test it with a mixed-fleet manifest. Set up three vehicle classes with different capacities, create runs for each and confirm that capacity, pricing and driver assignment follow the class. Then check that a full van stops selling seats while a coach on the same route stays open. Ask what happens when a larger vehicle is swapped in at short notice: do the booked passengers move across with their seats intact?

If you need to buy vehicles instead of managing your own, treat that as a sourcing question. Our guide to enterprise shuttle program platforms explains marketplaces, contractors and operator software. Shuttle.Software manages the operator side.

• Vehicle classes: each with its own seat count and rules.
• Capacity: a full vehicle stops selling; other classes stay open.
• Swap: a larger vehicle replaces a smaller one without losing seat assignments.
• Reporting: utilisation shown per class so you can right-size the fleet.

How do you test an AI route optimization claim?

Treat "AI routing" as a claim to test, not a feature to tick. The useful question is not whether a vendor uses AI, but whether the routes it produces are shorter, on time and explainable for your own bookings.

Run this test. Give each vendor the same 20 real bookings from a busy day, with their addresses, party sizes and hard arrival times. Ask for the sequence, total distance, each pickup time and any booking that cannot fit. Compare against what your dispatcher built by hand. A good system flags the infeasible booking before you accept it and lets a human override the result. Shuttle.Software sequences stops, times pickups backwards from the deadline and warns about infeasible runs; the dispatcher can always override.

• Same inputs: identical bookings for every vendor.
• Measure: total distance, lateness and the number of flagged problems.
• Explainability: can you see why a stop was ordered that way?
• Override: can a dispatcher change it and keep passenger times correct?
• Live tracking: does the passenger map and ETA update when traffic changes?

Does a no-code platform really work without technical skills?

Sometimes, for the changes it was designed for, and not for everything. "No-code" usually means you can change settings, fields, fares and screens without a developer. It rarely means any custom logic is possible. We have not tested other vendors' no-code tools, so we do not rate them here; test any claim yourself.

Ask the vendor to make one change while you watch, with a non-technical member of your team at the keyboard: add a booking field, change a fare rule and add a new pickup zone. Time it. If it needs a ticket, a developer or a quote, it is not no-code for your purposes. On Shuttle.Software, tailored routes, fares and booking fields are part of the Fleet plan; ask us to demonstrate the same three changes.

Where do the other questions on this topic go?

Several specific questions already have their own guides, so we link rather than repeat them. For sourcing vehicle age, seatbelt and amenity limits, see how small operators win enterprise contracts. For large programs, see the enterprise buyer's checklist. For white-label and custom app decisions, see custom taxi app vs white-label software. For a plain definition, see what is shuttle booking software.

Frequently asked questions

Keep reading