How to build an app like Urban Company: The hyperlocal services playbook

Urban Company's pitch sounds like TaskRabbit: book a home service, a trained professional shows up, pays through the app.

The difference is in the word "trained." TaskRabbit connects you with freelancers who already have skills. Urban Company trains professionals to deliver services their way -- standardized, audited, consistent. A "salon at home - full package" from Urban Company in Delhi and in Mumbai follows the same checklist, uses the same products, takes the same amount of time.

That standardization is the product. The app is how you deliver it.

If you're building in the hyperlocal services space, the technical architecture is similar to TaskRabbit. But the operational architecture -- how you build, train, and maintain professional supply quality -- is what makes the product work or fail.

Urban Company's model vs. TaskRabbit

Understanding the two models helps you decide which fits your market.

TaskRabbit model

• Customer browses professional profiles, picks one
• Open-ended or hourly pricing
• Professional sets their own rates
• Platform provides marketplace and payments
• Trust through verified reviews and background checks

Urban Company model

• Customer picks a service package (fixed scope, fixed price)
• Platform auto-assigns the best professional for that slot
• Platform sets prices (professionals earn a share)
• Trust through standardized service + training certificates + in-service audits

The Urban Company model requires more from the platform (training, service design, quality audits) but delivers more predictable quality at scale. The auto-assignment also means faster booking -- no browsing through profiles.

Which model fits your market depends on:

• Service standardizability: Can you define exactly what "deep home cleaning" means? Yes? Urban Company model works. "Handyman help" with undefined scope? TaskRabbit model is safer.
• Professional supply maturity: Does your market have professional tradespeople with established skills, or do you need to train supply from scratch? Established trades (electricians, plumbers) favor TaskRabbit. Services requiring platform-specific standards (beauty, wellness) favor Urban Company's training model.
• Customer decision-making: Do customers want to pick their professional (personal trust, recurring relationships), or do they want a reliable service without the browsing friction?

Customer app: What you're building

Service catalog and booking

Unlike food delivery, where the catalog is dynamic (restaurant-dependent), Urban Company's service catalog is platform-defined. "Salon at Home - Full Bridal Package" is a specific thing with a specific price, scope, and duration -- regardless of which professional delivers it.

This simplifies the booking experience:

1. Customer selects service category (beauty, cleaning, AC repair, etc.)
2. Customer selects specific service package
3. Customer picks date, time slot, and address
4. Platform assigns professional, customer gets confirmation

No browsing profiles. No price negotiation. No "what's included?" ambiguity.

The service catalog is a CMS: service categories, service packages with descriptions and prices, duration estimates, included materials/products (if any), and FAQs. Building and maintaining this catalog is operational work -- but the tech implementation is a structured content store, not complex logic.

Pre-service communication

After assignment, the customer sees the professional's name, photo, rating, and years of experience. They can message the professional through in-app messaging for pre-service questions (parking, access codes, special requests).

This is a trust-building moment. The professional was auto-assigned, so the customer hasn't "chosen" them. Surfacing the professional's profile -- even after the fact -- builds confidence.

Live tracking

When the professional is en route, the customer app shows live location on a map plus estimated arrival time. This reduces "where is the professional?" support queries, which are among the highest-volume support categories for home services platforms.

Post-service review

After service completion, the customer rates on multiple dimensions: quality, punctuality, professionalism, cleanliness. These dimensional ratings feed both the professional's quality score and the platform's audit queue -- a professional with consistently low "cleanliness" scores for a cleaning service gets flagged for re-training, not just a general low rating.

Ask for a review within 2 hours of service completion (while experience is fresh). Automated reminder if they don't review within 4 hours.

Professional app: The supply side

This is where most platforms underinvest. The professional app isn't just a job list -- it's the professional's entire work management tool.

Onboarding and certification

When a professional applies, the app walks them through:

1. Profile setup (name, photo, service categories)
2. Document upload (identity proof, address proof, skill certificates if applicable)
3. Background check submission
4. Training module completion (platform-specific service standards)
5. Practical assessment (for skilled services -- beauty, electrical -- a test booking with a quality assessor)

Training modules are video + quiz format, delivered in the app. A professional in the "Salon at Home" category watches 6-8 training videos on platform standards (hygiene, product application, customer interaction) before their first booking. This takes 2-4 hours.

Training module infrastructure: This is content delivery -- video hosting (Cloudinary or Mux), progress tracking, quiz scoring, and certification state management. Budget $10K-$20K for this layer.

Job feed and schedule management

Active professionals set their working days, working hours, and service area (city zone level). When a customer booking matches their skills, zone, and availability, the system checks their calendar. If there's a free slot, they're in the eligible pool.

Unlike TaskRabbit where professionals "accept" individual jobs, Urban Company's auto-assignment means professionals are assigned without explicit confirmation -- they receive a notification and are expected to show up. This requires professionals to keep their availability calendar current.

The job feed shows upcoming assignments: service type, customer name, address, time, and estimated earnings. Calendar view shows their week at a glance.

In-service checklist

This is unique to the Urban Company model and is the quality control mechanism. When a professional starts a service, they open a checklist in the app:

• Salon at home - full package: Check off: setup area cleaned and prepped, products within expiry, client consultation completed, service steps 1 through N followed, post-service product recommendations shared.
• Deep cleaning: Check off: rooms covered, surfaces cleaned in order, before-photo uploaded, after-photo uploaded.

Checklists serve two purposes: they guide the professional (especially newer ones) and they create an audit trail. The platform can review whether checklists were completed and whether photos were uploaded -- even without being present.

Photo documentation -- before and after service -- is required for some service types. These are reviewed by quality operations and can flag problems (damage not reported, service shortcuts) and also recognize excellent work.

Earnings and incentives

Professionals see per-job earnings, weekly totals, and payout history. Clear earnings visibility is a retention tool -- professionals who see consistent, growing earnings stay on the platform.

Urban Company uses a "partner kit" model: professionals pay a subsidized fee for branded products and uniforms, which are deducted from earnings. This creates product consistency (all professionals use the same cleaning products, beauty products) and a mild retention mechanism.

Incentive programs -- bonus per booking during peak hours, bonus for maintaining a 4.8+ rating, referral bonuses for recruiting new professionals -- drive supply quality and volume. These are configurable in the admin platform.

Matching engine: Auto-assignment logic

Urban Company auto-assigns professionals. The matching engine runs at booking confirmation to select the best professional for the job.

Eligibility filter

• Offers the service category requested
• Zone covers the customer's address
• Has an open slot at the requested time
• Background check cleared
• Training certification current for service type
• Quality score above platform floor (e.g., 4.0+)

Ranking the eligible pool

• Quality score (primary rank factor)
• Proximity (secondary -- closer professionals mean shorter travel time and lower lateness risk)
• Utilization (prefer professionals with fewer bookings that day -- don't overload top performers)
• Response to platform metrics (professionals with high cancellation rates deprioritized)

Capacity management

Professionals have a maximum jobs-per-day limit (set by them and enforced by the platform). A deep cleaning that takes 4 hours can't be double-stacked with another 4-hour job. The matching engine needs to check job duration against available time windows, not just "is there a free slot?"

Quality control system

Quality control is Urban Company's competitive moat. Building it requires four system components:

1. Service checklists: In-app task lists during service delivery. Built into the professional app as described above.

2. Photo audits: Before/after photo upload during certain service types. A quality ops team (or eventually an AI model) reviews photo submissions and flags quality issues.

3. Dimensional ratings: Post-service ratings from customers across 4-5 quality dimensions. Dimensional ratings surface specific problems -- a professional consistently rated low on "punctuality" needs a different intervention than one rated low on "skill quality."

4. Automated re-training triggers: When a professional's quality score falls below threshold on a specific dimension, the system enqueues them for a re-training module on that topic. In-app notification, module completion required before taking new bookings in that category.

5. Quality score model: A rolling weighted average of customer ratings, checklist completion rates, photo upload compliance, and cancellation rate. This score feeds the matching engine (high-quality professionals get more bookings) and identifies professionals at risk of deactivation.

Payment model

Urban Company's payment model differs from TaskRabbit's hourly model.

Customer payment: Fixed price per service package. No tipping historically (though they've experimented with optional tips). Credit/debit cards, UPI, digital wallets at checkout.

Professional earnings: Platform commission-based. Urban Company takes 15-25% of the booking value. The professional earns 75-85%. The exact split varies by service category and city.

Payment flow: Customer pays at booking (or post-service for some categories). Platform holds funds. After service completion with no dispute raised within 24 hours, professional payout processes weekly.

Equipment/product kit deductions: If the platform supplies branded products (cleaning supplies, beauty products), the cost is deducted from the professional's payout. This is tracked per-service and deducted at payout.

Technical stack

Customer app: React Native (iOS + Android). Mapbox or Google Maps for professional tracking.

Professional app: React Native. Offline-capable job viewing (professionals work in homes with sometimes unreliable connectivity). Photo upload with retry logic for poor connections.

Backend: Node.js or Python (FastAPI). PostgreSQL for transactional data. Redis for availability caching. Elasticsearch for service catalog search. Firebase Realtime Database or WebSockets for live professional tracking.

Training modules: Video hosted on Cloudinary or Mux (adaptive streaming). Custom LMS module in the professional app (quiz logic, progress tracking, certification state).

Payments: Razorpay in India (supports UPI natively). Stripe for global markets.

Background checks: Checkr (US/Canada), AuthBridge (India), Certn (Canada/UK).

Development timeline

Cost summary

The supply problem is not a tech problem

You can build a technically excellent hyperlocal services app and have nothing to show customers on day one if you haven't built professional supply.

Urban Company solved this through Pro Centers -- physical training and certification hubs in each city. Professionals come in for training, leave certified, and start taking bookings. The supply acquisition funnel (beauty school graduates, electricians looking for steady work, former hotel staff) is an HR and marketing function, not a tech function.

At launch scale, you don't need Pro Centers. You need:

• 20-30 trained professionals per launch city, per service category
• A training program (even if video-based) they've completed before their first booking
• A mechanism to collect and respond to quality issues fast (the first 30 days of bookings will surface process gaps you didn't anticipate)

The code ships. The supply takes longer. Start supply recruitment 3 months before you plan to launch.

1Raft builds marketplace and on-demand platforms for teams who understand this distinction. If you're planning a hyperlocal services product, talk to us before you start building -- scoping gets a lot more accurate when you've seen how supply-demand dynamics play out across similar products.