Case Study: Building Custom Software for an Auto Glass Business in Dallas

Auto glass repair looks like a straightforward service business from the outside: customer calls, technician comes, glass gets fixed. The operational reality is considerably more complex — and the complexity is where local businesses either build competitive advantages or get eaten alive by national chains with more resources.

The auto glass business in DFW specifically has layers that make generic service business software a poor fit. Insurance coordination is the most significant: the majority of auto glass jobs are insurance claims, which means the business is dealing with multiple insurance carriers, each with different approval processes, different pricing schedules, and different documentation requirements. Add to that the complexity of mobile service (technicians are dispatched to customer locations, not to a shop), real-time scheduling that accounts for parts availability, and the customer experience expectations of DFW's demanding consumer market, and you have a business that off-the-shelf field service software simply doesn't serve well.

Here's what we built for an auto glass business in Dallas, and why each component was built the way it was.

The Problem: Operations Running on Manual Bridges

When we started the engagement, the business was handling all insurance coordination manually: a dispatcher would call insurance companies, navigate hold trees, document claim numbers in a spreadsheet, and relay information to technicians via text message. The process worked, barely — but it was costing approximately 25 hours per week of dispatcher time, had an error rate that was generating roughly two disputed claims per month (each requiring hours of remediation), and created a customer experience that was opaque and anxiety-inducing for customers who had no visibility into where their claim stood.

The scheduling system was a combination of a scheduling tool and manual phone calls. Parts availability was tracked in a separate spreadsheet. Technician assignments were made based on dispatcher knowledge of who was available and approximately where, without systematic routing optimization. The business was doing respectable volume on this manual infrastructure, but the ceiling was visible — growing volume would require growing the dispatcher headcount linearly, and the error rate would grow with volume.

What We Built: An Integrated Operational Platform

Rather than automating pieces of the existing process, we designed a system that restructured the workflow around what digital tools can do that humans can't.

Insurance integration layer: the system connects directly with the major insurance carriers' API networks — where they exist — and uses structured data processing with AI assistance for the carriers that still require phone or portal-based interaction. When a job comes in with an insurance claim, the system identifies the carrier, pulls up the carrier's pricing schedule and documentation requirements, pre-populates the claim fields from the job record, and queues the claim for submission. For the handful of carriers with API access, claims are submitted automatically and status is updated in real time. For the majority that require web portal interaction, the system generates a structured checklist with all required fields pre-filled so the dispatcher interaction is reduced from twenty minutes of navigation to three minutes of copy-paste and confirmation.

Dispatch and routing optimization: technician assignments are now made by a routing engine that considers current technician location (pulled from a GPS integration with the technician's mobile app), estimated job duration based on job type and vehicle, parts availability at the nearest warehouse, and customer requested time window. The dispatcher reviews a ranked list of recommended assignments rather than building assignments from scratch. Override is always available, but the recommended assignment is right enough often enough that dispatcher cognitive load has dropped substantially.

Customer communication system: the most impactful user-facing improvement was a real-time communication system that gives customers visibility they didn't have before. When a claim is submitted, the customer gets an automated notification. When the claim is approved, another notification. When a technician is assigned, the customer gets the technician's name, estimated arrival time, and a link to a web view that shows real-time technician location as they drive. On job completion, an automated satisfaction check and review request sequence initiates.

Technician mobile app: the technician side of the system handles job acceptance, navigation integration, parts confirmation, photo documentation (before and after), and job completion with customer signature. The photos and signature attach automatically to the job record, which feeds directly into the invoice and insurance documentation workflow.

The Outcomes

The measurable outcomes twelve months after deployment are worth documenting specifically, because they illustrate what the right software investment actually delivers.

Dispatcher labor for insurance coordination dropped from 25 hours per week to 8 hours per week — a 68% reduction. The two disputed claims per month dropped to less than one every two months — an 80% reduction. Both improvements came from the same root cause: structured data handling replaced manual data transcription, eliminating the error rate that transcription carries.

Customer satisfaction scores improved significantly. The primary driver of negative reviews before the system was customer anxiety about claim status — not knowing whether the claim was approved, when the technician would arrive, and whether the job was done correctly. The real-time communication system addressed each of these directly. Review volume also increased because the post-job review request sequence has a 22% response rate versus the 4% response rate of the previous manual follow-up process.

Scheduling efficiency improved enough that the same dispatcher headcount handles 35% more daily volume. This created growth capacity without proportional overhead growth — exactly the kind of operational leverage that makes a software investment compelling.

What This Illustrates About Custom Software for Service Businesses

The auto glass example is instructive because none of the individual components we built were technically exotic. Insurance integration, routing optimization, customer communication, mobile apps for field technicians — each of these exist as components in various commercial products. What doesn't exist is a product that combines them in the specific way this business needs, with the specific carrier relationships and pricing logic baked in, with the specific workflow that reflects how this team actually operates.

That's the case for custom software in service businesses: not that commercial products are bad, but that they're built for the median business. The median auto glass business is not the same as this business — it doesn't have the same insurance carrier mix, the same parts supplier relationships, the same customer segment expectations. Software built for the median serves the median. Software built for your specific business serves you.

The investment required to build this system was substantial. The return — measured in dispatcher labor savings, error reduction, customer satisfaction improvement, and growth capacity — produced a positive ROI inside eighteen months. The software continues to evolve as the business evolves, which is what Living Software means in practice.

If you run a service business in DFW and recognize your operational problems in this description, the conversation worth having is about what a system built specifically for your business would change. That conversation starts at routiine.io/contact.