Solutions · Mobility & logistics

Taxi booking app: what to decide, build and test.

Planning one? Here's what to settle before you brief anyone, what goes in the first version, how the build runs week by week, how the beta is tested with real people, and the numbers that show it's working.

Four decisions that change the whole taxi build

Four choices decide what gets built and which licences you need. They are cheaper to argue about on paper than in code.

The first column is a question for you. The last column is why it is not a technical question.
DecisionOption AOption BThe trade-off
Who owns the vehiclesYour own fleet or leased cars with salaried or contracted driversOpen marketplace of drivers with their own vehiclesA fleet gives you control of quality and a simple payout, but you carry the vehicles, the idle time and the insurance. A marketplace scales faster, but you need driver onboarding, checks and a payout split you can defend.
Service shapeOn-demand, matched in secondsScheduled: carpool, airport, rental, intercityOn-demand needs live matching and a dense driver pool. Scheduled needs seat inventory and reminders, and works with a handful of drivers.
Money flowFare goes through you, then payout to the driverRider pays the driver directly, you bill a subscription or feeTaking the fare gives you control and data, and also puts GST and payout logic on you. Direct payment is lighter, and gives you less to enforce.

Our concept build Hoppr took one path: bike taxi first, marketplace captains, fare through the platform, one city rule set per city. That is one answer, not the answer. The taxi app guide goes deeper.

Before you brief anyone, be able to answer these:

  • Which city launches first, and what is the written rule for your vehicle type there?
  • Do you hold, or plan to apply for, an aggregator licence, or will you operate under a licensed partner?
  • Does the fare pass through you, or straight to the driver?
  • Who answers an SOS alert at 11 pm, and what do they do?

Version 1 of a ride app: what ships and what waits

A launch in one city needs far less than a large platform's feature list, but some pieces cannot be skipped.

ModuleFirst versionAdd laterWhy
Rider bookingPickup, drop, vehicle choice, fare, driver card, OTP startScheduled rides, saved routes, multi-stopBooking and OTP are the loop. Extras do not move trust.
Driver appDocuments upload, duty toggle, request accept, navigation hand-off, earningsIncentive streaks, in-app trainingDrivers must be able to onboard without visiting you.
MatchingNearest available driver, one retry radiusBatching, demand predictionSimple matching works until you have hundreds of drivers.
Fare engineBase, per km, per minute, minimum, with a fixed surge ceilingZone-level surge, passes, couponsFare rules are where legal limits live, so keep them few and visible.
PaymentsUPI intent or QR, cash, receiptWallet, saved cards, corporate billingCash and UPI cover most Indian riders at launch.
SafetySOS button, live trip share, masked callingRoute-deviation alerts, audio recordingSOS is the part that cannot wait. Deviation alerts can.
Ops boardLive map, drivers online, active trips, SOS queue, document review queueHeatmaps, forecasting, fraud rulesYour team runs the launch from this screen.

How a 10 to 12 week taxi build runs, and what you supply each week

The slowest items are usually yours: merchant account approval, licences and store accounts (see the Play Store launch guide). Start them in week 1.
WeeksWhat we buildWhat you supply
1–2Flows, fare rules written down, clickable prototype of rider and driver appsFare card for launch city, vehicle types, cancellation and refund policy, a few real drivers willing to test
3–5Rider app, driver app and backend, onboarding and document upload, map and place searchMaps and SMS accounts in your name, legal name and brand assets, list of documents you want drivers to submit
6–8Matching, live trip, OTP, fares, UPI and cash, SOS, ops boardPayment gateway merchant account (KYC takes time, start early), a named safety contact, test phones including low-end Android
9–10Driver review queue, notifications, website, test rides in the fieldTen to twenty drivers and riders for a controlled pilot, support phone number, privacy policy and terms from your lawyer
11–12Fixes from the pilot, store listings, launch support, handover of code and accountsPlay Store and App Store developer accounts in your name, final approval, launch-day staffing

How we work with you each week

The same rhythm runs from week 1 to launch, so you always know what's happening and what we need from you.

  • Friday: a demo link and a 30-minute call. You open the latest build on your own phone and try it yourself, rather than watching our screen.
  • Within 48 hours: your feedback, in writing. In one shared doc, next to the screen it's about. Written feedback is clear, dated and doesn't get lost between calls.
  • Monday: triage. Every comment goes into fix now, next week or later. Anything that changes the agreed scope gets a written quote before we touch it.
  • Each milestone: two revision rounds, then your sign-off. Sign-off is against the scope we agreed, so it's clear what done means.
  • One decision-maker on your side. Most delays aren't build time. They're feedback waiting on several people who disagree.

The full version, including how the beta and user testing work across every kind of project, is on our process page.

The beta: a controlled pilot in one zone

The pilot runs in weeks 9–10, once a ride can go from booking to OTP start to drop and payment without anyone on the ops board stepping in. It covers one zone of the launch city and one vehicle type.

  • What's in it. Booking with fare shown upfront, nearest-driver matching, OTP start, UPI and cash, the receipt, SOS with live trip share, masked calling, the driver app with earnings, and the ops board with its document review queue.
  • What stays out. Scheduled rides, coupons, zone-level surge, a wallet and corporate billing. Surge stays switched off so the pilot shows the base fare working first.
  • Who uses it. Ten to twenty people in all: six to ten drivers who already drive this vehicle type in the zone, on their own Android phones, and riders who make regular trips there, such as office commuters and college students. You recruit both. Friends riding around for fun don't book at 9 am in traffic or argue a fare, and that is what you need to see.
  • How they get it. Riders install through TestFlight or Google Play testing. Drivers get the Android build. Your named safety contact and support staff use the ops board on a staging link.
  • How long. One to two weeks of real rides and real fares, limited to set hours, such as morning and evening peaks, so ops can watch every trip. We run one planned SOS drill during the pilot.

A new personal Google Play developer account has to run a closed test with at least 12 testers for 14 days before it can publish. Count the pilot drivers and riders towards that, or use an organisation account.

What we test with real people, and the tasks we set

Testers get a task, not a questionnaire. They use their own phone, set to their own language, and we watch quietly without pointing at the screen. Five or so sessions for each role is usually enough to find what breaks. We fix it, then test again with fresh people.

The SOS drill is planned with your safety contact in advance, and riders are told it is a test.
WhoThe taskWhat we watch for
Rider"Book a ride from the metro station exit to your office, check the fare before you confirm, and pay by UPI at the end."Do they set the right pickup gate, not just the station? Do they understand the fare before booking? Does the UPI hand-off return to the app?
Rider"Share your live trip with a family member, then find the SOS button without looking for it in advance."Can they share in two taps? Can they find SOS quickly while the vehicle is moving?
Driver"Upload your licence and RC, go on duty, accept the next request and start the trip with the rider's OTP."Do photos of documents upload on weak data? Is the accept button easy to hit while stopped at a signal? Does the OTP step slow the start?
Driver"The rider isn't at the pickup point and isn't answering. Cancel with the right reason."Do they find the valid reasons? Do they understand whether a penalty applies before they cancel?
Safety contact"An SOS alert has just come in at 10:40 pm. Find the trip, call the rider and log what you did."How long until someone acts? Does the board show location, driver and vehicle number together?
Ops reviewer"Five drivers uploaded documents today. Approve the good ones and send one back with a reason."Can they tell a blurred RC from a clear one? Does the driver see the reason in their app?

The numbers that say it's working

You choose two or three of these in week 1. If you already take bookings by phone or WhatsApp, we write down today's figure to compare against. Tracking is built before the pilot, and we report at launch and 30 days later. The figures are not something we can promise. What we promise is that you can see them.

GoalWhat we measureWhere it comes from
Riders get a vehicleShare of booking requests that end in an accepted ride, by hourMatching logs
Pickups are quickTime from accept to driver arriving at pickupDriver app location and timestamps
Fares don't surprise anyoneFare disputes and complaints per 100 tripsOps board and support log
Drivers stay on the appDrivers who went on duty in the pilot and are still going on duty a month laterDriver duty data
Safety alerts are answeredTime from SOS to first response by your teamSOS queue
Cancellations stay explainableCancellations by side and by reasonTrip data

What a taxi app costs, and what moves it

A rider app, driver app and live ops board with maps and payments sits in the marketplace-with-live-maps-and-payments tier of our app development cost guide: a ballpark of ₹25–60 lakh at small-studio rates. If you run your own fleet and your own drivers you move toward the two-user-types-plus-admin tier of ₹10–25 lakh. These are ballparks, not a quote; the studio quotes after reading your brief.

  • Pushes it up: live location at short intervals for every online driver, because it raises map, server and battery work. Surge zones and extra cities add rules to test.
  • Pushes it up: automated document checks instead of a person reviewing uploads.
  • Pushes it down: a single city, one vehicle type and cash plus UPI only.
  • Not in the build price: maps, SMS, gateway fees, hosting and insurance. Ask for a monthly estimate.

Rules and habits to design for in India

Ride apps are among the most regulated app categories. The central Motor Vehicle Aggregator Guidelines, 2025 set the frame, and states adopt or adjust them, so the rule for your city and vehicle type is the one that counts. Hoppr stores each city's rules as data, so a rule change is an edit and not a rewrite (its Mumbai electric-only bike-taxi rule is our reading when built and needs legal review).

  • Surge: the 2025 central guidelines allow dynamic pricing up to two times the base fare, with discounts down to half of it. Some states have drafted lower caps, so build the ceiling as a setting per city.
  • Driver share: the guidelines set a floor of 80% of the fare for drivers who own their vehicle. Your payout logic has to prove it.
  • Insurance: the guidelines ask aggregators to arrange health cover of at least ₹5 lakh and term cover of at least ₹10 lakh for each driver. Budget it as a running cost.
  • Cancellations: a driver who cancels an accepted trip without a valid listed reason can be penalised up to 10% of the fare, capped at ₹100. The valid reasons must be published in the app and on the site.
  • Gig-worker contributions: under the Code on Social Security, aggregators contribute 1–2% of annual turnover, capped at 5% of what they pay gig and platform workers. Ask your CA when it applies to you.
  • GST: passenger transport booked through an app can make the platform responsible for the tax, and the rate options have changed more than once. Have your CA fix the treatment before you design the receipt.
  • Data: trip history, live location and phone numbers are personal data. The DPDP Rules were notified in November 2025 with duties phased in over 18 months, so build consent and deletion now. See our DPDP checklist.

This is general information, not legal advice. Rules differ by state and change often; confirm licences, fares, insurance and tax with a lawyer and a chartered accountant before launch.

Mistakes that sink new ride apps, and what to do instead

  • Launching in three cities at once. Each has its own fares, rules and driver pool. Pick one zone, get pickups fast, then expand.
  • Building rider features before driver supply. Start recruiting in week 1 and ship the sign-up page in version 1.
  • Treating SOS as a button. Name the person, the hours and the escalation, and test it before launch.
  • Hard-coding fares and surge. Rules change by city and by year. Keep them in settings your team can edit with a date.
  • Testing only on new phones. Drivers use older Android handsets and weak data, so test on those.
  • Leaving verification vague. Say which documents, who reviews them and how long it takes, and show the status in the app.

Cost and timeline

10–12 weeks for a first version.

Fixed quote within 48 hours of your brief: scope, timeline and price in writing before we start. These are what move the number.

  • Type of service: on-demand rides, scheduled carpool or rentals
  • Fare logic: per km, surge, shared fares or deposits
  • Live map tracking and how often locations update
  • Driver onboarding and document verification steps
  • Number of cities and languages at launch

How app costs work in India →

Questions

About your taxi booking app.

How much does it cost to build a taxi app in India?

It depends on the service type, fare logic, live tracking and how much driver onboarding you need. Every project gets a fixed quote within 48 hours of your brief, so you know the number before we start. See the app cost guide for what moves it.

Can you build a bike taxi or auto app, not just cabs?

Yes. Our concept build Hoppr is bike-taxi first, with autos as a second option, a captain app with a helmet check, and a live ops board. The same base works for cabs, carpools and rentals. You can open it and try every side.

How long does it take?

A full rider app, driver app, ops board and website usually takes 10 to 12 weeks. A rental or carpool product with simpler matching can be shorter. A two-week concept sprint gets you a clickable prototype to show investors or drivers first.

Will the driver app work on low-cost Android phones?

Yes. Drivers often use older handsets with limited data, so we keep the app light, cache what we can and test on low-end Android phones. Location updates are tuned to save battery during long shifts. Trip requests still arrive when the app is in the background.

Who owns the code and the brand?

You do. The source code, design files and brand system are handed over in full. Your team, or any other developer, can keep building on them. Driver records, trip history and fares stay in your own accounts, so nothing depends on us after launch.

Start a project

Tell us how your business runs. We'll build what it runs on.

Four taps, about a minute. Fixed quote within 48 hours.

Menu

Read

Ask an expert Ask an expertFree, real peopleHub
Ask an expert