Ride-Hailing · Delivery · Home Services · Healthcare · Super-App
On-Demand App Development Services — Three-Role Marketplaces Built to Dispatch, Not to Demo
Uber-style ride-hailing, Urban Company-style home services, Dunzo-style hyperlocal delivery, doctor-on-demand, tutor-on-demand, and salon-at-home builds — customer app, service-provider app, and admin dashboard delivered as one system. Priced fixed in INR, built from Ahmedabad, shipped for on-demand founders and enterprise operators across India, the Gulf, the UK, and the US since 2014.
- ✓Customer app, service-provider app, and admin dashboard shipped as one connected system
- ✓Live GPS tracking, distance-based pricing, surge, and pickup / drop geofencing wired in day one
- ✓Razorpay, Cashfree, PayU, Stripe, UPI, wallets, and cash-on-service checkout in one flow
- ✓Firebase push, WhatsApp Business API, SMS OTP, and deep links out of the box
- ✓App Store, Play Store, and Huawei AppGallery submission handled end-to-end by our team
Why an on-demand app is a different build from a standard mobile app
An on-demand product is not one app. It is a live three-sided marketplace running on a phone, held together by real-time dispatch and settled every 24 hours. The customer wants a two-tap request, a real-time ETA, and a price they can trust before they confirm. The service provider — driver, technician, doctor, tutor, beautician, or delivery agent — wants a job feed that respects distance, skill, and rating, a route that respects one-way traffic, and a payout screen that reconciles daily to the bank account. The operator on the admin side wants live coverage of every job, every provider, and every payment, with a support console that can intervene inside 60 seconds when a customer or a provider raises a red flag.
That is the wall a generic template collapses at. A boilerplate app from a marketplace like CodeCanyon buys a ten-week head start and then two years of firefighting when the dispatch engine breaks under load — jobs stuck in "assigned" for hours, providers pushed to the wrong pickup, refunds that never process, and Play Store reviews that read "app not working". A no-code tool like Bubble or Glide will get a founder to a demo, but the moment real dispatch, real surge, and real settlement enter the picture, the ceiling is right there.
Fruxinfo has been shipping custom mobile apps from Ahmedabad since 2012 and on-demand marketplace builds since 2016. What we ship today is a proper three-role system — customer, provider, admin — on one shared backend, wired to real payment rails, real maps, and real messaging, priced fixed in INR up-front, and handed over with source code you own from day one. The apps go through App Store and Play Store review with our team on the paperwork, not left as a checklist for the founder.
MVP scope for an on-demand app — the minimum that actually ships
Every founder who calls us about an on-demand app arrives with a feature list that reads like Uber's roadmap in year six. The right first move is the opposite — cut the scope to what a real customer can request from on launch day, what a real provider can accept and complete, and what a real operator can supervise from the admin. Everything else is a version-two problem.
The MVP we ship for a single-city launch is a working three-role system with the modules below. It goes live in twelve to sixteen weeks and lets you take real requests from real customers the day the app hits the store — not a beta that needs six more months of polish. The scope is fixed, the price is fixed, and every change request is estimated in writing before work starts.
- ›Customer app (Flutter iOS + Android) — phone OTP sign-up, category picker, service or vehicle selection, address / pickup picker, price preview, request, live tracking, chat, ratings
- ›Service-provider app (Flutter Android, optional iOS) — sign-in, availability toggle, incoming job card with 15-second accept window, navigation, in-progress screen, complete flow, daily payout view
- ›Admin dashboard (web) — live job map, provider onboarding with KYC upload, category and pricing engine, coupon and referral engine, settlement console, dispute queue
- ›Backend and API — Node.js (NestJS / Express) or Laravel with a WebSocket layer for real-time job events and a Redis-backed dispatch queue
- ›Payments — Razorpay or Cashfree online checkout, cash-on-service flow with provider reconciliation, refund-to-source in 3-5 working days
- ›Location and routing — Google Maps SDK for customer live tracking, Directions API for provider navigation, geofencing on pickup and drop for status verification
- ›Notifications — Firebase FCM + APNS for push, WhatsApp Business API for booking confirmation and ETA nudges, SMS fallback for OTP and critical status changes
- ›Compliance — DPDP Act 2023 consent capture, GST invoicing per job, RBI card tokenisation, App Tracking Transparency copy for iOS 14+, data-deletion route
Cost to build an on-demand app in India — the real INR ranges
On-demand app pricing is the single most-asked question on the first call, and the honest answer is: it depends on how much of the three-role system you actually need on day one and how many verticals you are running. Below are the ranges we quote after a proper discovery — not a range pulled from a marketing blog. Every number assumes a Flutter build (one codebase, iOS + Android), a Node.js or Laravel backend, source code in your Git, and App Store plus Play Store submission handled by us.
A single-vertical booking app — one service category, customer app plus provider app plus a lightweight admin, no live dispatch (the provider just picks a slot from a queue) — starts at around ₹3 lakh and ships in eight to twelve weeks. This is the shape most home-service, salon-at-home, tutor-booking, and doctor-appointment apps launch on.
A live-dispatch on-demand app — customer, provider, and admin, with real-time job assignment, GPS tracking, and dynamic pricing — sits in the ₹6-14 lakh range and ships in twelve to sixteen weeks. Ride-hailing (Uber / Ola-style), on-demand delivery (Dunzo / Porter-style), and technician-dispatch products for AC repair, plumbing, and electrical work live in this bracket.
A multi-vertical hyperlocal super-app — one customer app that lets a user request rides, food, groceries, and home services from a single home feed, with separate provider apps per vertical and a unified admin — runs ₹15-35 lakh and ships in twenty to thirty weeks. This is the shape a well-funded regional operator picks when the goal is one wallet, one login, and a shared referral engine across all services.
A full Uber-style multi-city marketplace with surge pricing, subscription plans, ads infrastructure, driver leasing, cash-collection reconciliation across cities, and a serious ops console is quoted after discovery and starts at ₹30 lakh, with a six to nine-month timeline for the first production city and a copy-and-scale cadence for city-two onwards.
Recurring costs are honest and small. Google Maps runs about USD 200-900 per month at launch scale, climbing with active-user growth (Places, Directions, and Distance Matrix APIs are the meters that move). Firebase for push and analytics is free for most launch traffic and rarely crosses USD 100 per month before the fifty-thousand-MAU mark. Razorpay charges 2 percent on domestic cards and UPI is free for merchants under the current MDR rules. WhatsApp Business API conversation cost runs ₹0.85-2.85 per notification. App Store fees are USD 99 per year for Apple, USD 25 one-time for Google — accounts opened under your business name.
- ›Single-vertical booking app (no live dispatch) — ₹3-6 lakh, 8-12 weeks
- ›Live-dispatch on-demand app (ride / delivery / technician) — ₹6-14 lakh, 12-16 weeks
- ›Multi-vertical hyperlocal super-app — ₹15-35 lakh, 20-30 weeks
- ›Uber-style multi-city marketplace — from ₹30 lakh, quoted after discovery
- ›Add-on vertical on an existing super-app — ₹4-10 lakh per new category
- ›iOS + Android via one Flutter codebase — no separate native quotation unless the brand insists
- ›App Store USD 99/year + Play Store USD 25 one-time — accounts opened in your name
- ›Real running cost (Maps + Firebase + Razorpay MDR + WhatsApp) — ₹18,000-70,000/month at launch scale
Feature list by role — what actually gets built in each app
The three-role system is not three copies of the same screen with the colour swapped. Each app does one job for one user, and the value of the build is in how tightly they talk to each other. Below is the module-by-module feature list we ship on a standard live-dispatch on-demand build — the reference scope every on-demand quote starts from.
Customer app is where the visible product lives. Sign-in runs on phone OTP with Google and Apple sign-in as backup. Home screen shows the service categories (or vehicle classes, for ride-hailing) with icons, and a search bar wired to a Meilisearch or Typesense index for typo-tolerant lookup. Service selection carries variants, add-ons, and a price preview computed off distance, time-of-day, and category. Address picker uses Google Places autocomplete with a "confirm on map" pin. Booking screen supports "book now" and "schedule for later". Live tracking is a Google Maps view with the provider's marker, ETA, and a call-provider button (masked number via Exotel or Knowlarity). Order history, saved addresses, saved cards (tokenised via the payment gateway), and referral share close the loop.
Service-provider app runs on the provider's Android phone, tuned for cheap devices and patchy 4G. Availability toggle, incoming job card with a 15-second accept window, pickup navigation, in-progress screen (with time-tracker for hourly services or checklist for salon / repair jobs), complete flow with photo upload and customer signature capture, cash-collection reminder, and end-of-day payout view. Every screen is one-tap, high-contrast, and works with the phone locked in a bike mount. Optional voice-navigation cues in Hindi, Gujarati, Marathi, or Tamil for drivers who do not read English fluently.
Admin dashboard is the operator's control room. Live job map of every open request across the city. Provider onboarding with Aadhaar via DigiLocker (where possible), driving licence, vehicle documents, skill certificates for regulated categories, and background-check status. Category and pricing engine with per-category base fare, per-km rate, per-minute rate, night surcharge, and holiday surcharge. Coupon engine with stackable rules, first-order caps, and category-specific promos. Payment and settlement reconciliation against the Razorpay settlement report. A refund console with source-back-to-card support. A dispute queue for jobs where customer or provider flags a problem, with a chat log, GPS breadcrumb, and payment history stitched together on one screen.
A support ops console is the fourth screen every serious operator adds around month three — a manual-dispatch overrider, a live-support chat inbox, a bulk-notification composer, and a suspended-provider dashboard. We build this at launch on any project above the ₹15 lakh mark.
Tech stack and integrations a live on-demand product runs on
An on-demand product lives on the reliability of about ten integrations. Miss one and the app fails the launch-week fire drill. Below is the reference stack we ship, and the integrations Indian on-demand founders actually ask for on the first call. Every line item plugs in through the vendor's official API, so nothing is a scraped hack that breaks at the next vendor release.
The frontend is Flutter for iOS and Android (one codebase, native performance) on both the customer and provider apps, with Next.js for the admin dashboard on the web. The backend runs on Node.js (NestJS or Express) or Laravel, depending on the team you plan to grow later. Realtime job state is a WebSocket layer (Socket.IO or Laravel Reverb) with a Redis pub-sub channel underneath, so a status change on the provider app reaches the customer's live-tracking screen in under a second. The dispatch engine itself is a Redis-sorted-set queue with a geohash-based provider lookup — assignments happen in under 200 milliseconds even at a few thousand active providers per city.
Storage is PostgreSQL or MySQL for transactional data, Redis for live state and dispatch, and S3-compatible object storage for provider documents and job proof photos. Push runs on Firebase Cloud Messaging plus Apple Push Notification Service, wired through a single messaging service so a WhatsApp fallback fires automatically when push fails.
Payments plug into Razorpay, Cashfree, PayU, PhonePe, and Paytm — one primary gateway with an optional secondary fallback so a gateway outage does not stop bookings. UPI, cards, netbanking, wallets, and EMI are surfaced in one checkout. Cash-on-service is handled with a provider-side reconciliation flow and a daily settlement report. Refunds fire back to the original payment method inside 3-5 working days. For cross-border on-demand builds (Gulf, UK, US), Stripe and Adyen are the primary rails.
Maps and location run on Google Maps SDK for customer tracking and the Directions API for provider routing. Mapbox is an option when the Google bill starts to hurt at scale. Geofencing on pickup and drop points is standard — the provider has to be inside the geofence before a status can move to "picked up" or "completed", cutting fake status changes. Distance Matrix API prices jobs dynamically based on real driving distance, not straight-line kilometres.
Compliance and messaging round out the stack. GST invoicing on every job with GSTIN, HSN codes, and place-of-supply logic. WhatsApp Business API through AiSensy, Interakt, Gupshup, or Wati for booking confirmation, ETA nudges, and reviews. SMS OTP via MSG91, Textlocal, or 2Factor. Number masking via Exotel or Knowlarity so customer and provider phone numbers stay private. Analytics wires GA4, Meta Pixel + CAPI, Firebase Analytics, and either Mixpanel or Amplitude for funnel and retention reporting. DPDP Act 2023 consent capture, RBI card-tokenisation compliance, and a data-deletion route are built at launch, not retrofitted.
- ›Flutter iOS + Android (customer + provider apps), Next.js web admin
- ›Node.js (NestJS / Express) or Laravel backend with Socket.IO or Laravel Reverb realtime
- ›Redis-backed geohash dispatch engine — sub-200ms provider assignment at scale
- ›PostgreSQL or MySQL + Redis + S3 — boring, battle-tested data stack
- ›Razorpay, Cashfree, PayU, PhonePe, Paytm — dual-gateway fallback available
- ›Google Maps SDK + Directions + Distance Matrix, or Mapbox — with pickup / drop geofencing
- ›Firebase FCM + APNS push, WhatsApp Business API, SMS via MSG91 / Textlocal / 2Factor
- ›Exotel / Knowlarity number masking, GST invoicing, DPDP consent, RBI tokenisation — compliance built in
Monetization models — how on-demand apps actually make money in India
An on-demand app has more revenue levers than most first-time founders realise, and picking the right mix at launch protects the unit economics for the next two years. Below are the models we architect for, with real India-market ranges. Most successful builds use two or three of these in combination, not just one.
Commission per job is the standard marketplace model. Uber and Ola charge drivers 20-25 percent per ride, Urban Company charges professionals 18-25 percent per job, and Dunzo takes a per-delivery fee plus a merchant commission. A new city-level marketplace typically starts at 12-18 percent to win supply, then walks the number up as density grows. Commission is deducted before settlement and appears on the provider's daily payout view.
Convenience fee to the customer is the second lever. Standard range is ₹10-45 per booking depending on category and time of day, with surge pricing during peak hours or bad weather. The pricing engine we ship supports flat, distance-based, and surge tiers, and lets the operator experiment on the admin panel without a code deploy.
Subscription plans are the retention lever. A ₹99-499 monthly plan that waives convenience fees, adds priority dispatch, and layers in category-specific perks (free ride credits, discounted salon visits, unlimited grocery delivery). We wire subscription as a Razorpay Subscriptions or Stripe Billing product with a customer-facing manage-plan screen. Subscription revenue is the highest-margin line item once the base grows past a few thousand active customers.
Provider advertising and priority placements are a mature-market lever. Sponsored positions at the top of a category feed, boosted profiles for verified providers, and priority job assignment for providers on a paid tier. We ship the placement infrastructure at MVP and turn it on when supply density justifies a sales conversation with providers.
Cross-selling across categories is the super-app lever. A customer who booked a ride today gets a nudge for a home-service tomorrow. A user who ordered groceries this week sees a laundry offer next week. The wallet, referral engine, and loyalty points are shared across categories — the customer acquisition cost is amortised across every service.
Card-linked cashback partnerships, first-booking coupons funded by providers or brands, referral credits, and small transaction fees on UPI (when MDR returns) round out the mix. The right combination depends on the city, the price point, and the supply density — we walk through the P&L math on the blueprint call before the build starts.
- ›Commission per job — 12-25% of booking value, deducted before provider payout
- ›Convenience or platform fee to customer — ₹10-45 flat or surge-priced at peaks
- ›Subscription plan — ₹99-499 monthly, waived fees + priority dispatch + credits
- ›Provider advertising and priority placements — sponsored profiles, boosted jobs
- ›Cross-category upsell in a super-app — shared wallet, referral, and loyalty points
- ›Cash-collection margin — small float retained on cash bookings before settlement
- ›Corporate accounts and B2B billing — monthly-invoiced enterprise clients
- ›Card-linked cashback, first-booking coupons, referral credits — retention levers
On-demand verticals we build for — where the pattern actually fits
Every on-demand enquiry that lands on our contact form fits into one of the vertical shapes below. The core three-role system is the same across all of them — customer app, provider app, admin — but the pricing engine, the job-life-cycle screens, and the compliance modules change by category. Knowing your vertical lets the first call move quickly to the right reference and the right number.
Ride-hailing and taxi apps — Uber, Ola, InDrive-style. Vehicle-class selection, per-km + per-minute pricing, surge, driver-side rating gates, and RTO document verification. Optional intercity, rentals, and outstation modules. Typical build ₹8-16 lakh, live-dispatch grade.
On-demand delivery and courier apps — Dunzo, Porter, Shadowfax-style. Pickup / drop with a category picker (documents, groceries, food, parcels), COD support, weight-and-size-based pricing, and multi-stop routing. Typical build ₹7-14 lakh.
Home services and technician dispatch — Urban Company, Housejoy-style. Category catalogue (AC repair, plumbing, electrical, appliance service, pest control), slot-based booking or live dispatch, technician kit management, before-and-after photo capture, and warranty tracking. Typical build ₹8-16 lakh.
Beauty and salon-at-home — Yes Madam, GlamOnGo-style. Multi-service basket per visit, professional-level filtering (senior stylist vs junior), package deals, membership plans, and hygiene-kit compliance checks. Typical build ₹6-12 lakh.
Healthcare on-demand — Practo, DocsApp-style teleconsult plus at-home care. Doctor discovery by specialty and rating, video consult (integrated over Agora, Twilio, or Jitsi), e-prescription with a digitally-signed PDF, ABDM (Ayushman Bharat Digital Mission) integration for the Health ID and health-records flow, and pharmacy-refill deep link. Typical build ₹10-22 lakh for the on-demand slice alone.
Tutoring, coaching, and skill-on-demand — Vedantu, Unacademy-live-style. Tutor discovery, one-to-one or group session booking, in-app video, whiteboard, session recording, and progress tracking. Typical build ₹9-18 lakh.
Laundry, cleaning, and daily-help — UrbanClap-lite / Broomberg-style. Recurring subscriptions, per-visit or per-load pricing, pickup + drop scheduling, and a chore checklist. Typical build ₹5-11 lakh.
Pet care, fitness, and specialised on-demand niches — dog-walking, in-home yoga, personal trainer, at-home fitness class, pet grooming. Same three-role shape, per-visit or subscription pricing, and either a live-dispatch or a scheduled-slot flow. Typical build ₹5-12 lakh.
Roadside assistance, on-demand mechanics, and emergency services. SOS button on the customer app, nearest-provider dispatch, ETA guarantees, and integration with the fleet insurance or roadside-assistance policy. Typical build ₹8-16 lakh.
If your vertical is not on this list, that is normal — most real projects sit at the boundary between two. The blueprint call is where we place the project on the map and quote against the closest reference.
Timeline and delivery process — how we ship an on-demand build
We work in fixed-scope, milestone-billed sprints with a weekly demo. The clients we ship for know what will be on staging every Friday, and they know exactly which invoice covers which milestone. Below is the standard timeline for a live-dispatch on-demand build — the most common shape we ship in the ₹6-14 lakh range.
Weeks 1-2 are the blueprint phase. On-site or Zoom discovery with the founder, ops lead, and (if the client is a service business) the head of field operations. We walk the current job flow — how a WhatsApp or phone booking arrives today, how the provider gets assigned, how the money moves back — and we come out with a written scope document, wireframes for all three apps, a data model, a payment-and-refund flow, a dispatch-algorithm spec, and a fixed-price quotation. Nothing gets coded until the scope is signed.
Weeks 3-6 are backend and admin. The backend, database, admin dashboard, provider onboarding, category and pricing engine, and Razorpay integration are the first modules built and demoed. The admin panel goes live on staging by end of week 6, and pilot providers can start entering documents and testing the sign-up flow.
Weeks 7-11 are the customer app. The Flutter customer app is built module by module — onboarding, home, category, address, booking, live tracking, order history, ratings. Each module hits a weekly Friday demo. Google Maps, push, deep links, and payments wire in as horizontal capabilities. By end of week 11 the customer app is on internal-test tracks on TestFlight and Play Internal Testing.
Weeks 12-15 are the provider app and the launch hardening. Provider sign-in, availability, incoming-job card, navigation, in-progress flow, complete flow, and cash reconciliation. End-to-end soak tests with real providers in a live cluster. App Store and Play Store review submission on week 13. Launch checklist — DPDP Act consent audit, GST invoice sample audit, RBI tokenisation audit, App Store and Play Store review reply — signed off before go-live.
Post-launch, the first month is a hyper-care sprint — daily standup, live production monitoring on Sentry and Datadog, and same-day patches on P0 issues. From month two, the app moves to the standard retainer cadence — a fortnightly release, a monthly analytics review, and a quarterly roadmap conversation.
The next step — book an on-demand app blueprint call
The fastest way to know what your on-demand app should actually cost and what should be in scope is a 30-minute conversation. We will ask about the vertical (ride, delivery, home service, healthcare, tutoring, salon, laundry, or something else), the city, the launch job volume you expect, the payment mix your customers prefer, and the timeline you are working against. At the end of the call you get a written blueprint — recommended scope, module list, integrations, indicative timeline, and a fixed-price quotation — with no obligation to move forward.
Call +91-99245-12890 or drop an enquiry through the contact form. We reply the same working day.
On-demand app questions we hear on the first call
What ride-hailing, hyperlocal delivery, home-service, and healthcare on-demand founders ask before starting a build. If yours isn’t here, we will answer it on the discovery call.
Explore Related Services
Combine these services with what you just explored for a complete digital strategy from Fruxinfo.
Mobile App Development
Native and cross-platform iOS, Android, and Flutter apps for modern business needs.
Learn more →Food Delivery App Development India
Custom Zomato/Swiggy-style food delivery apps — customer, restaurant, and delivery-partner apps on one system, with Petpooja POS, Razorpay, live GPS tracking built in.
Learn more →Native App Development Services
Native iOS (Swift, SwiftUI) and Android (Kotlin, Jetpack Compose) builds for hardware-heavy, latency-critical, and OS-day-one apps — fixed-scope, source-code ownership.
Learn more →
