Good field service app development starts with one assumption: the technician will lose signal at the worst possible moment. Build offline-first, track location without draining the battery, and model your real dispatch rules before you design a single screen. Get those three right and the rest is ordinary app work.

This guide covers what we build for home services, utilities and logistics teams. It walks through the features that matter and the regulations that change the build. It also covers typical integrations, cost and timeline ranges, and the mistakes we see most often.

Five features field service app development has to get right

1. Offline-first data and sync

Basements, rural lines and steel warehouses kill connectivity. The app must read and write everything locally, then sync when a connection returns. That means a local database such as SQLite or WatermelonDB, or a sync engine like PowerSync. A cache bolted onto API calls is not enough.

The hard part is conflicts. Say a dispatcher reassigns a job while the tech is offline and closing it. Someone's change has to win, and the app must decide which one.

We set a rule per field. Last write wins for notes. The server wins for job assignment. Photos and time entries are append-only, so nothing a tech captured is ever overwritten.

2. Battery-aware GPS tracking

Dispatch needs to know where trucks are. Techs need a phone that lasts a 10-hour shift. Continuous high-accuracy GPS will not give you both.

We use high accuracy only while a tech is driving to a job. On site, the app falls back to low-power updates. A distance filter of 50 to 100 meters cuts updates sharply without hurting the dispatch map.

Platform rules matter here. On Android 14 and later, background tracking needs a foreground service declared with the location type. On iOS, you must justify background location during App Store review, as Apple's Core Location documentation describes.

3. Dispatching and scheduling

The dispatch board is where the office lives all day. It must account for skills, certifications, time windows, parts on the truck and drive time. Drag-and-drop is the minimum, while suggested assignments save the most time.

Route planning is a well-studied problem with mature tools. Google's open-source OR-Tools handles vehicle routing with time windows. Google Maps Platform or Mapbox add live traffic. Start with suggestions a dispatcher can override, not full automation.

4. Forms, photos and signatures that hold up later

Paper forms become digital checklists, but they must work offline and capture evidence. Every completed job should store photos, a customer signature, a timestamp and a GPS stamp. That record settles warranty disputes and insurance claims.

Compress photos on the device to roughly 200 to 500 KB before upload. Otherwise, a tech who shoots 40 full-size images can wait minutes on a weak connection. Queue uploads in the background so the next job is never blocked.

5. Live job status and customer updates

Customers want an arrival window and an on-my-way text. Status changes from the tech should send those messages automatically: en route, on site, complete. In home services, this typically cuts where-is-my-tech calls more than any other feature.

You get sharper answers to your own build questions when you talk to an engineer who has built in your industry.

Regulations that change the build

This is not legal advice. Which rules apply depends on your trade, your vehicles and your states. These are the ones that most often change what we build.

Logistics: FMCSA hours of service and ELDs

If your drivers operate commercial motor vehicles, the FMCSA electronic logging device rule under 49 CFR Part 395 applies. Do not build your own ELD, because devices must be registered with FMCSA. Integrate with a registered device through your telematics provider instead.

Utilities: OSHA and NERC CIP

OSHA's recordkeeping rule, 29 CFR 1904, shapes how injury and incident forms store data. Lockout/tagout under 29 CFR 1910.147 often becomes a required checklist step before work starts. If the app touches bulk electric system assets, NERC CIP standards drive access control and audit logging.

Payments and customer data: PCI DSS, TCPA and state privacy laws

Taking cards in the field puts you in PCI DSS scope. A certified reader through Stripe Terminal or Square keeps raw card data off your servers and shrinks that scope.

Automated customer texts fall under the TCPA, so capture consent at booking. Employee location history is personal data under laws like California's CCPA. Set a retention period and tell workers exactly what you track.

Integrations you should budget for

Most field service apps are a hub between systems you already run. Integration work typically takes 25 to 40 percent of the build budget. Expect most of these:

  • Accounting: QuickBooks Online, Xero or NetSuite for invoices and payments.
  • CRM: Salesforce or HubSpot for customer and contract data.
  • Asset and work management: IBM Maximo, SAP or Oracle, common in utilities.
  • GIS: Esri ArcGIS for poles, valves, meters and network maps.
  • Telematics: Samsara or Geotab for vehicle location and driver logs.
  • Messaging and maps: Twilio for SMS, plus Google Maps Platform or Mapbox for maps and routing.

Ask each vendor about API rate limits early. A legacy ERP that allows only a few requests per second can force a nightly batch design. That is far cheaper to plan for than to discover in month four.

Buy, extend or build: the honest comparison

Off-the-shelf: ServiceTitan, Jobber, Housecall Pro

These are the fastest way to launch, often within weeks. They fit standard home services workflows well. Per-tech subscription fees grow with your team, and the vendor controls your roadmap.

Platform extension: Salesforce Field Service, Dynamics 365 Field Service

This makes sense if you already run that CRM. You pay licences plus configuration work. Unusual offline needs or complex compliance steps can still require custom development.

Custom build

This has the highest upfront cost, but you own the workflow, the data and the roadmap. It pays off when your scheduling rules, compliance steps or integrations are a real edge. It also pays off when licence fees for 100 or more techs exceed the build cost over three to five years.

Realistic timeline and cost

These are typical ranges for a US-based team building iOS, Android and a web dispatch console. Your scope will move them in either direction.

  • Discovery and workflow mapping: 3 to 5 weeks, $15,000 to $35,000.
  • MVP with offline jobs, forms, basic dispatch and one or two integrations: 4 to 6 months, $120,000 to $200,000.
  • Full product with route optimization, payments, a customer portal and four or more integrations: 6 to 9 months, $200,000 to $350,000.
  • Utility-grade with GIS, asset management and regulated data: often 9 to 12 months or more, from $350,000.

Plan 15 to 20 percent of the build cost each year for maintenance, OS updates and integration changes. A cross-platform framework like React Native or Flutter often saves 25 to 35 percent versus two native apps. That holds for this kind of form-heavy, data-driven product.

Three mistakes first-time buyers make

Treating offline as a phase-two feature

Offline support changes the data model, the API and the sync logic. Adding it later typically means rewriting the core of the app. Decide on day one, even if the first release only covers a few screens offline.

Designing in the office instead of the truck

Techs wear gloves, stand in bright sunlight and hold a ladder with one hand. Tap targets need to be at least 44 by 44 points, contrast must be high, and common actions should take two taps. Ride along with three techs before you approve designs.

Digitizing paper instead of fixing the process

A six-page paper form becomes a six-page digital form nobody finishes. Cut fields the office never reads, prefill from the job record, and make the rest conditional. Then pilot with 5 to 10 techs for two to four weeks before full rollout.

Two questions to ask any agency

Can you walk me through a sync conflict in an app you shipped?

You want a specific answer: which data won, why, and what the user saw. Vague talk about real-time sync usually means they have not built it. Ask how they recovered when sync failed in production.

How do you test with bad or no network?

Good answers name tools and methods. Look for Network Link Conditioner on iOS, Android emulator throttling, scripted airplane-mode tests and field testing on real devices. If testing means the office Wi-Fi, expect bugs in the basement.

FAQ

How much does it cost to build a field service app?

An MVP typically costs $120,000 to $200,000 and takes four to six months. A full product with payments, route optimization and several integrations usually runs $200,000 to $350,000. Budget another 15 to 20 percent of the build cost each year for maintenance.

Can a field service app work without internet?

Yes, if it is built offline-first with a local database on the device. Techs can view jobs, fill forms, take photos and capture signatures with no signal. Everything syncs when a connection returns, using conflict rules you define up front.

Should I buy ServiceTitan or build a custom app?

Buy if your workflow matches standard home services and you want to launch in weeks. Build when your scheduling rules, compliance steps or integrations do not fit the product. Building also makes sense when per-tech fees at your scale outgrow the cost of owning the software.