A plumber I know runs 6 trucks in northern New Jersey. His office manager starts the day with 8-12 voicemails from overnight. Half are emergencies ("my basement is flooding"). Half are service calls ("kitchen faucet is dripping"). She triages them, checks which tech is closest, calls the tech, calls the customer back, confirms the time window, and moves on to the next one. By 9:30 AM she's still processing the morning queue, and the phone is ringing with new calls.
His best tech spent 45 minutes driving past two other jobs to reach a call that any of his guys could have handled. Nobody caught it because dispatch was doing everything in their head.
I build AI systems for service businesses. Trades companies are a natural fit because dispatch, invoicing, and follow-up are all pattern-based work that happens at high volume. Five automations below.
The problem: A service call comes in. The dispatcher mentally runs through: which tech is available, who's closest, do they have the right skills for this job type, do they have the parts on their truck, and can they fit it in before their next appointment? With 6 trucks, that's manageable. With 10+, mistakes happen. The wrong tech gets sent. The drive time is longer than it needs to be. A specialized job goes to a generalist, and they can't finish it in one visit.
The automation: An AI agent assigns incoming calls to technicians based on GPS location, skill set, current workload, truck inventory, and job complexity. A residential panel upgrade goes to the tech with electrical panel experience who's 8 minutes away, not the apprentice who's 3 minutes away but would need to call for backup. A simple faucet replacement goes to whoever's closest and has the most common faucet parts on their truck.
The agent also detects scheduling conflicts. "Tech Mike is assigned a 2 PM water heater install in Montclair (estimated 3 hours) and a 4 PM service call in Hackensack. The 4 PM won't work. Reassigning to Tech Dave who finishes at 3:30 PM in Paramus."
One company reduced average drive time between jobs from 28 minutes to 16 minutes. That's an extra job per tech per day. For 6 trucks at an average ticket of $350, that's $2,100 per day in additional capacity.
Additional capacity: 1 extra job per tech per day. Setup cost: $2,500-$5,000. Monthly cost: $150-$300.
The problem: The tech finishes a job. He writes up the invoice on a paper form or taps it into a tablet. He drives to the next job. The invoice goes to the office. The office enters it into QuickBooks. The office sends it to the customer. The customer pays in 15-45 days. Or doesn't, and the office chases them with phone calls. Average collection time: 23 days. Cash flow suffers.
The automation: An AI agent generates the invoice the moment the tech marks the job complete. It pulls the line items from the work order, applies the correct pricing, adds materials used, calculates tax, and sends the invoice to the customer via text and email with a pay-now button. The customer pays on their phone while the tech is still in the driveway.
For customers who don't pay within 3 days, the agent sends a reminder. Day 7: another reminder. Day 14: phone call prompt to the office with the full account history. Day 30: final notice before collections.
One company reduced average collection time from 23 days to 4.7 days. That's 18 days of improved cash flow per invoice. For a company billing $80,000/month, that's about $48,000 in cash that used to be floating and is now in the bank.
Collection time reduction: 60-80%. Setup cost: $1,500-$3,000. Monthly cost: $80-$150.
The problem: A homeowner calls at 11 PM because water is pouring from a burst pipe. Your after-hours answering service takes a message. You see it at 6 AM. The homeowner already called three other plumbers and whoever answered first got the job. Emergency calls are the highest-revenue tickets in the trades ($500-$2,000+), and most companies lose them because they can't respond fast enough outside business hours.
The automation: An AI agent handles after-hours calls. It asks structured triage questions. "Is there active flooding? Is the water near electrical panels or appliances? Have you turned off the main water valve?" Based on the answers, it categorizes urgency and takes action.
True emergency (active flooding, gas smell, sparking wires): the agent immediately texts the on-call tech with the address, customer name, and triage details. It also texts the customer: "Our tech Mike is on his way. Expected arrival: 35 minutes. In the meantime, please turn off the main water valve. It's usually near your water meter."
Urgent but not emergency (no hot water, toilet won't flush, outlet not working): the agent books a first-available morning slot and confirms with the customer. "We'll have someone there between 8-9 AM tomorrow. You'll get a text when the tech is on the way."
Non-urgent (dripping faucet, flickering light): the agent books a regular appointment and sends confirmation.
One company captured 8-10 additional emergency calls per month that they were previously losing to competitors with faster response times. At $800 average emergency ticket, that's $6,400-$8,000/month in captured revenue.
Emergency calls captured: 8-12 additional per month. Setup cost: $2,000-$3,500. Monthly cost: $100-$200.
The problem: A plumber installs a water heater with a 6-year warranty. Six years later, nobody remembers, and the customer calls when it fails catastrophically. An electrician installs a panel and recommends an inspection in 5 years. Nobody follows up. Every completed job is a future maintenance opportunity, and most trades companies leave that revenue on the table because nobody tracks it.
The automation: An AI agent tracks every installed system, warranty date, and recommended maintenance schedule. Water heater installed in 2026? The agent sends a maintenance reminder in 2028. "Hi Tom, it's been two years since we installed your water heater. A quick flush and anode rod inspection keeps it running efficiently and protects your warranty. Want to schedule a maintenance visit? $125 for the full inspection."
The agent also sends seasonal reminders tied to your service area. "Freezing temperatures expected next week in Bergen County. Exposed pipes in crawl spaces and garages are at risk. Want us to add pipe insulation? Quick job, $150-$250 depending on accessibility."
Post-job follow-ups go out 48 hours after every service call. "Hi Sarah, this is a follow-up on the outlet repair Tuesday. Everything working properly? If you notice any issues, reply here and we'll come back at no charge." The happy customers are primed for a review request. The unhappy ones tell you before they post on Google.
Repeat service revenue: $3,000-$8,000 per month from maintenance reminders. Setup cost: $1,500-$3,000. Monthly cost: $60-$120.
The problem: Reviews are the #1 driver of new customers for trades companies. But techs don't ask for them (they feel awkward), and office staff don't follow up consistently. Your Google profile has 47 reviews from 3 years of work, and the competitor down the road has 200 because they bought a review management tool. Except their tool sends a generic email 3 days later that most people ignore.
The automation: An AI agent sends a review request 2 hours after job completion, via text (not email, because text open rates are 95% versus 20% for email). "Hi Tom, thanks for choosing us today. If the work went well, would you mind leaving a quick review? It takes 30 seconds." Link goes directly to your Google review page, pre-opened to the 5-star button.
The timing matters. Two hours after the job, the customer is still thinking about the service. Two days later, they've moved on. The text message format matters. It feels personal, not automated.
For customers who rated 4-5 on the follow-up survey, the review request goes out automatically. For customers who rated 3 or below, the agent routes them to the office for resolution before asking for a review. You don't want unhappy customers leaving public reviews.
One company went from 3-4 new reviews per month to 18-22 per month. Within six months they had 150+ reviews and ranked #1 in their local Google Maps pack for "plumber near me." They told me new calls increased 35% from Google alone.
Reviews per month: 4-5x increase. Setup cost: $1,000-$2,000. Monthly cost: $40-$80.
Total setup for all five automations: $8,500-$16,500. Monthly running cost: $430-$850. Run your numbers. Between extra jobs from better dispatch, faster cash collection, captured emergency calls, maintenance revenue, and new customers from reviews, most trades companies see $15,000-$25,000 in additional monthly revenue within the first quarter.
Emergency call routing first if you're losing after-hours calls. It's the highest revenue per automation dollar spent.
Review generation first if your Google presence is weak. It's the cheapest to set up and has the longest compounding effect.
I've written about why AI projects fail. Trades companies sometimes get sold on expensive field service management platforms ($300-$500/month per tech) that promise AI dispatch. Most of those AI features are basic routing. Build focused automations that solve your specific problems for a fraction of the cost.
Take the AI readiness quiz to see where your company is losing the most money.
Your customers don't care if an AI or a receptionist confirmed their 2 PM appointment. They care that someone answered the phone when their basement flooded at midnight.
Running a trades company and losing calls, cash flow, and reviews? Let's fix that.
Book a Free Strategy Call