For Dispatchers

Pineido resources for dispatchers in HVAC, plumbing, electrical, and home services. SOPs, KPIs, FSM patterns, and the toolkit we use on every dispatch sprint.

5 min read6 sectionsApril 26, 2026

Written by

Kostadin Stamboliev

Founder and Imaginator

Founder and Imaginator at Pineido. Practitioner across IT consulting, app and web development, technical SEO, and AI / automation. Ships production code in 30-day sprints with fixed scope and clear ownership.

LinkedIn

Daily routine that works

Dispatcher arrives 90 minutes pre-shift. 30 minutes for plan, 30 minutes for tech check-in, 30 minutes for board lock. In-shift 30-minute pulse to review the board. Post-shift 60 minutes to close the day and stage tomorrow. The 90+60 minutes around the shift is what enables real dispatch, not reactive dispatch. Best-practice shops also plan the board a week or two out, then adjust daily, rather than building each day from scratch.

KPIs dispatchers should own

On-time arrival rate (target 90 percent; optimized shops with clean routing hit roughly 91 percent versus a 76 percent baseline). First-time fix rate at the shop level (target 80 percent or higher; this is the single best signal of dispatch and assignment quality, and real HVAC turnarounds have moved FTF from 62 percent to 84 percent with better truck stock and skill matching). Calls per dispatcher per day (capacity check). Re-dispatches per day (friction signal). Day-end FSM data integrity (every job closed in software).

FSM workflows to set up

Skill-based and territory-based assignment is free in the core dispatch board of ServiceTitan, Jobber, and Housecall Pro; set up skills and business units so the board only surfaces techs who can actually do the job. ML auto-assign is a paid tier: ServiceTitan Dispatch Pro scores techs on skill, drive time, and historical close rate, recommends an assignment, and lets the dispatcher accept, reject, or override. The Fall 2025 board rebuild added a job holding area for parking unscheduled work, and Smart Routing rolls out through summer 2026. Beyond assignment: standard templates for emergency vs standard dispatch, required fields enforced before a tech can close a job, and ETA/reminder automations to protect SLAs. Most FSMs support this; few shops configure it.

When to push back on owner

Dispatcher should push back when: (a) Owner books outside capacity. (b) Owner skips intake protocol for VIP customer. (c) Owner overrides tech assignment without context. Empowering the dispatcher to push back protects the day.

Career path

Dispatcher, senior dispatcher, dispatch supervisor, operations manager. Entry HVAC dispatchers in 2026 average around $45K to $47K base (roughly $21 to $22 per hour), with experienced dispatchers reaching $60K to $70K. Senior, supervisor, and high-cost-metro roles, plus total comp with bonus and overtime, push the top of the range toward $110K. Dispatch is a real career, not an entry-level role.

How agentic AI takes this further

ServiceTitan's Dispatch Pro and Housecall Pro's assignment scoring are a step past plain automation, they weigh skill, drive time, and history and recommend a tech, but they still stop at recommend and wait for the dispatcher to click accept. A real agent closes that last step for the jobs where the recommendation is not actually in question: it perceives the new job, decides against the same rules a good dispatcher already carries in their head, and acts by making the assignment and firing the notification, only stopping short on the calls that genuinely need a person. That is different from a chatbot dispatch assistant that answers "who's free right now" in a sentence, which is a faster way to ask a question, not a system doing the dispatcher's job for them.

The agent watches the FSM's job queue continuously, each tech's live status and location where the FSM exposes GPS, and the skill and business-unit tags already configured on the dispatch board. Those are the right inputs because a dispatcher's mental model runs on exactly this data refreshed in real time, and an agent that only checks it once an hour will make the same stale-data mistakes a distracted dispatcher makes, just with more confidence.

Standard assignment logic, skill match, territory, current load, runs as explicit rules because that part of the job genuinely is deterministic once the data is live; an LLM call handles the harder calls, a VIP account that should go to a specific senior tech, a tie between two equally qualified techs, grounded with retrieval-augmented generation over the shop's own dispatch history and customer notes. Below the confidence threshold, an emergency call with no clean skill match, or a tech whose availability status has not updated recently enough to trust, the agent surfaces the case to the dispatcher instead of assigning on a guess.

Acting means a direct FSM API write to the job's assigned-tech field and a Twilio or native push notification to the tech, idempotent against the job ID so a webhook retry does not double-assign the same job to two techs. If the FSM returns an error, or shows the job already assigned by the time the write lands, a dispatcher who beat the agent to it, the run halts and defers to the human assignment rather than overwriting it, because overwriting a dispatcher's own call the one time it happens is enough to make them stop trusting the system.

The dispatcher keeps the checkpoint on exactly the cases the rules do not cleanly cover, and critically, keeps the authority to override anything the agent does, in real time, from the same board they already use. That authority is what makes this different from the black-box ML scoring already built into some FSMs: the dispatcher can see why the agent made a call and adjust the rule set directly, rather than trusting an opaque score with no visibility into its logic.

We build this on n8n against the client's own FSM API and Twilio account, on infrastructure the client owns, and hand the dispatcher the ability to see and adjust the rules directly rather than filing a ticket with us or another vendor every time a rule needs a tweak. That is a deliberate contrast with a rented dispatch-AI product, where the assignment logic is the vendor's intellectual property and the dispatcher has no path to understanding, let alone changing, why it made a given call.

What it will not do: replace the dispatcher's judgment on emergency calls, VIP accounts, or anything the rule set has not been taught to handle, those stay supervised because the cost of a wrong call there is real, and it will not run well on a shop whose FSM data is too dirty to trust, garbage skill tags produce garbage assignments regardless of how good the agent's logic is. The realistic outcome is not a black-box efficiency number; it is a dispatcher who spends the day managing exceptions and the board instead of retyping the same routine assignment forty times a shift. Pineido builds AI agents for small business into the dispatch workflow itself, one owned workflow live in 30 days, not a black box the dispatcher has to trust blindly.

Get operations insights weekly

Join service-business operators who get one actionable insight on ops, automation, and growth every week.

Checking that you're human. You can send in a moment.

No spam. Unsubscribe anytime.

You’re subscribed! Look for our next issue in your inbox.

Ready to ship?

Tell us where you are stuck. We will review it and send a focused recommendation within one business day.

Frequently asked questions

How many techs can one dispatcher handle?

8 to 12 active service techs in a residential shop. Headcount can read higher (10 to 15 on paper) when many techs run installs or projects rather than live service, so manage to active service load, not raw roster. Strong FSM discipline raises the ceiling. Past 12 active, add a second dispatcher with a territory split; past 24, add a dispatch supervisor.

Should dispatchers also do CSR work?

In sub-5-tech shops yes. Past 5 techs the roles diverge; merging them creates bottlenecks.

How does Pineido help dispatchers?

Dispatch optimization sprint includes dispatcher coaching, KPI dashboard, FSM workflow buildout, and 14 days of live shadow.