Every HVAC qualification call asks the same six questions. Whether a person asks them or an AI does is the only variable. The difference is what happens to the answers.
A CSR writes notes. Notes get summarized. Summaries get handed off. By the time the dispatcher sees what the lead actually said, the lead has called two more companies. An AI voice agent captures the same six answers as structured CRM fields the moment the call ends, and a workflow routes the lead to the right next step before the homeowner even hangs up.
TL;DR
- HVAC qualification has six criteria: service type, system status, equipment age, property type, service area, and timeline. Every one of them affects what the lead is worth and how fast to respond.
- The problem isn’t the questions. It’s that the answers disappear into call notes that nobody acts on automatically.
- The ServiceAgent AI voice agent conducts the qualification conversation on every inbound call and writes each answer as a structured CRM field. The Workflow Builder then reads those fields and routes the lead without anyone deciding who handles it.
- Three routing paths: qualified-high-urgency (immediate dispatch), qualified-routine (confirmation + standard queue), and disqualified (out-of-area or non-homeowner). Each path runs on its own node sequence.
Why Manual HVAC Qualification Loses the Data?
The qualification conversation itself is not where most HVAC operators lose leads. The questions get asked. The homeowner answers. The CSR listens.
The loss happens at handoff. The answers live in a call note, a notepad entry, or a verbal summary passed to the dispatcher. None of those formats trigger an action automatically.
The dispatcher reads through notes and decides who to call back, in what order, and with what urgency. On a slow Tuesday, that works fine. On a peak summer day when fifteen leads came in before noon, it doesn’t.
The structural problem: manual qualification separates the information from the action. The answers are captured in one place (the note) and acted on in another place (the dispatcher’s judgment). Every handoff between those two steps introduces delay, interpretation error, and dropped context.
Automating qualification means the AI captures the answers as structured data, and the workflow acts on that data without a handoff. The lead doesn’t wait for a human to read the note and decide what happens next.
The Six HVAC Qualification Criteria
These six criteria determine what every HVAC lead is worth and what the right response is. Each one needs to be captured on the first call, not inferred later.
Service type. Repair, replacement, maintenance, or new install. A repair call and a system replacement call have different ticket values, different urgency levels, and different dispatcher assignments. Capturing this on the call means the right tech gets the right job.
System status. Is the system running at all, running poorly, or fully down? A fully down system in July is an emergency. A system that’s cycling short but still producing some cooling is urgent but not an immediate dispatch. The status determines the response speed the workflow applies.
Equipment age. Units over 12 to 15 years old change the conversation from repair to replacement. An AI that captures equipment age can flag replacement-eligible calls for a different follow-up than a repair call on a three-year-old unit. That distinction is worth thousands of dollars in average ticket value.
Property type. Homeowner or renter. A renter typically can’t authorize HVAC work. Sending a truck to a renter who didn’t get landlord approval wastes a dispatch. Capturing this on the call filters it before the truck rolls.
Service area. Does the address fall within your service radius? An out-of-area call that makes it to the dispatcher’s queue burns time that should go to qualified leads. Automated service area verification on the call closes that gap at the intake point.
Timeline. Emergency (system down now), this week (degraded performance), or planning ahead (getting quotes, not yet buying). Timeline determines which routing path the lead enters and how much follow-up it needs if it doesn’t book on the first call.
All six criteria affect routing. None of them are useful sitting in a call note.
What “Automatic” Qualification Actually Means?
Most tools marketed as “automated lead qualification” for HVAC do one of two things. They send an SMS or email when a form is submitted (notification, not qualification). Or they use an IVR call flow where the homeowner presses 1 for emergency and 2 for maintenance (self-selection, not conversation).
Neither approach captures the six criteria above. SMS doesn’t ask anything. IVR captures one data point (urgency self-selection) and misses equipment age, system status, property type, and service area entirely.
True automatic qualification means an AI voice agent conducts the qualification conversation on the inbound call, in natural language, and writes every answer as a structured field in the CRM the moment the call ends. The homeowner answers questions the same way they would with a CSR. The difference is what happens next: the answers become data, and data triggers a workflow.
Introducing the Workflow Builder
The Workflow Builder is a visual drag-and-drop canvas inside ServiceAgent where you build automated sequences that run the moment a trigger event fires. Each workflow starts with a trigger (the system event that starts the sequence) and runs through a chain of nodes, which are individual actions the system takes automatically without staff involvement.
For HVAC qualification, the trigger is contact.created: it fires the moment the AI voice agent creates a new contact record from the inbound call. From that trigger, two nodes classify the lead using the qualification data, and three routing paths handle each outcome.
Book a 20-minute demo to see the Workflow Builder in action.
The Qualification Workflow: Two Nodes, Three Paths
This workflow fires the moment a new contact is created from an inbound HVAC call. It runs on every lead, at any hour, without staff involvement.
Trigger: contact.created
Configure the trigger by selecting “New Workflow” in the Workflow Builder and choosing contact.created as the event. Set the trigger to fire for all new contacts sourced from the AI voice agent. This workflow handles qualification routing. A separate confirmation workflow handles contacts that booked during the call.
What to check: create a test contact via the AI voice agent and confirm the workflow activates in the activity log. If it doesn’t fire, verify the source tag condition matches what the AI voice agent writes on new contact records.
Node 1: AI Extract (capture all six criteria)
What it does: Reads the full transcript from the AI voice agent call and extracts all six qualification criteria as structured CRM fields: service type, system status, equipment age, property type, service area verification, and timeline.
Why it matters: The AI voice agent conducts the qualification conversation during the call. AI Extract translates that conversation into structured data the Workflow Builder can act on. Without this node, the transcript stays as unstructured text. With it, each qualification answer becomes a field the next node reads to make a routing decision.
What happens: The node processes the call transcript and writes six values to the contact record. Service type (repair / replacement / maintenance / new install). System status (fully down / degraded / functional). Equipment age in years, or a flag for “age unknown” if the homeowner couldn’t answer. Property type (homeowner / renter / unknown). Service area (in-area / out-of-area, verified against your service ZIP codes). Timeline (emergency / this-week / planning-ahead).
What to check: After a test call, open the contact record and confirm all six fields are populated. If system status or property type is returning “unknown” consistently, the AI voice agent’s qualification prompt needs those questions added explicitly. A blank field is not the same as a known answer of “unknown.” Fix the extraction before taking the workflow live.
Node 2: AI Decision (route by qualification outcome)
What it does: Reads the six fields written by Node 1 and routes the workflow to one of three paths based on the combined qualification outcome: Path A for qualified high-urgency leads, Path B for qualified routine leads, and Path C for disqualified leads.
Why it matters: This is the routing decision that previously lived in the dispatcher’s judgment. The AI Decision node makes it automatically, consistently, and in under a second, based on the same criteria a CSR would use. It doesn’t replace the dispatcher’s expertise on the job itself. It removes the triage step that was eating the first 10 to 20 minutes of every busy morning.
What happens: The node evaluates the qualification fields and applies the routing logic you configure. High-urgency qualification: system status is “fully down” AND property type is “homeowner” AND service area is “in-area.” Routine qualification: service type is repair, maintenance, or replacement, system status is degraded or functional, property type is homeowner, service area is in-area. Disqualified: property type is “renter,” OR service area is “out-of-area,” OR timeline is “planning-ahead” with no clear service need.
What to check: Test all three paths explicitly. Run a call with a homeowner reporting a fully down system in-area. Run a call with a homeowner requesting a tune-up. Run a call from a renter. Confirm each routes to the correct path. Misroutes at this node send high-urgency leads to the wrong queue or block qualified leads from reaching the dispatcher.
Path A: Qualified High-Urgency (emergency dispatch sequence)
Path A runs when Node 2 identifies a lead as qualified and high-urgency: homeowner, in-area, system fully down. These leads need to reach the dispatcher immediately.
Node A1: Create Task (high-priority dispatch)
What it does: Creates a high-priority dispatch task in the CRM with all six qualification fields visible in the task body, routed to the on-call tech or dispatcher with an urgent flag.
Why it matters: A fully down system in July is a same-day job. The task surfaces at the top of the dispatcher’s queue with everything they need: what broke, how old the system is, what the homeowner’s address is, and that this lead is already confirmed as a homeowner in your service area. The dispatcher doesn’t need to read a transcript or call back to verify. The qualification already happened.
What happens: The task is created with urgency: high, due: today, and the qualification fields from Node 1 in the task body. Routing assigns it to the dispatcher or on-call tech based on the rules you set once during configuration.
What to check: Confirm the task appears at the top of the high-urgency queue in the dispatcher’s view. Verify all six qualification fields are visible in the task body. If the equipment age field is blank, the dispatcher needs to know that going in rather than seeing a gap they have to fill during the dispatch call.
Node A2: Send SMS (immediate confirmation)
What it does: Sends a confirmation SMS to the homeowner within 60 seconds of the contact being created, confirming their request was received and that someone is following up now.
Why it matters: A homeowner whose AC is down in July called you and two other companies. An SMS within 60 seconds signals responsiveness before the dispatcher has made the callback. It anchors the homeowner to your company while the dispatch is being assigned, reducing the chance they book with a competitor who happened to call back first.
What happens: “Hi [Name], this is [Business Name]. We received your [Service Type] request and someone from our team will be in touch shortly. Questions? Call or text [Number].” Sent within 60 seconds of the trigger.
What to check: Confirm the SMS arrives within 60 seconds on a test run. Verify the service type field populates correctly from Node 1.
Path B: Qualified Routine (standard queue sequence)
Path B runs for qualified leads that are not high-urgency: maintenance requests, planned replacements, degraded-but-functional systems, and routine repairs that don’t require same-day service.
Node B1: Send SMS (confirmation + booking prompt)
What it does: Sends a confirmation SMS within 60 seconds that includes the service type captured during the call and an offer to schedule.
Why it matters: Routine leads aren’t racing the clock the way emergency leads are, but they’re still comparing options. A same-day confirmation SMS that includes a next step (schedule now or call to discuss) keeps them in your pipeline without requiring dispatcher time to make an outbound call.
What happens: “Hi [Name], thanks for calling [Business Name] about your [Service Type]. We’d love to get you scheduled. Reply here to book or call [Number] and we’ll find a time that works.” The SMS populates from Node 1 fields.
What to check: Confirm the message arrives within 60 seconds. Check that the service type populates correctly.
Node B2: Create Task (standard queue, next-day)
What it does: Creates a standard-priority task in the dispatcher queue with the qualification fields, due the next business day.
Why it matters: Routine leads don’t need same-day dispatch, but they do need a queue entry the dispatcher can work through systematically. Without this node, routine leads accumulate in a contact list with no priority signal, and the dispatcher decides who to call based on whoever they remember or whoever called most recently.
What happens: Task created with urgency: standard, due: next business day, and all six qualification fields in the task body. Equipment age appears here as well: if the unit is 14 years old and the homeowner is inquiring about repair, the dispatcher can initiate the replacement conversation during their callback rather than after an unnecessary repair visit.
What to check: Confirm the task appears in the standard queue with the correct due date and that equipment age and timeline fields are visible.
Path C: Disqualified (polite close sequence)
Path C runs for leads that don’t qualify for service: out-of-area callers, renters without landlord authorization, or planning-ahead calls with no near-term service need.
This path is the one no competitor in this SERP covers. Most articles say “route disqualified leads to voicemail.” That’s a dead end for the homeowner and a wasted touchpoint for your business. A polite close sequence acknowledges the lead, explains why you can’t serve them, and leaves a better impression than voicemail.
Node C1: Send SMS (polite close)
What it does: Sends a brief SMS that acknowledges the call and explains that you’re unable to help, either because the address is outside your service area or because the job requires homeowner authorization.
Why it matters: A renter who calls your company and hears nothing will tell their landlord about a bad experience with your brand. A renter who receives a polite “we’d need landlord authorization for this work” message understands the situation and is more likely to share your contact with the landlord. Out-of-area callers who get a referral (even a vague one) leave with a positive impression. Neither group becomes a dispatcher burden.
What happens: For out-of-area: “Hi [Name], thanks for calling [Business Name]. Unfortunately your address falls outside our current service area. We’d recommend checking [Google Maps/Angi/your local HVAC directory] for licensed contractors near you.” For renters: “Hi [Name], thanks for reaching out. HVAC work at a rental property typically requires authorization from the property owner or landlord. If you can get us in touch with them, we’d be glad to help.”
What to check: Confirm the SMS matches the disqualification reason from Node 2 (out-of-area vs. renter). A generic message that doesn’t explain the reason reads as dismissive.
Node C2: Update CRM (disqualified tag)
What it does: Tags the contact record as “Disqualified” with the specific reason (out-of-area or renter) and a timestamp, then closes the workflow.
Why it matters: Disqualified contacts that sit in your pipeline without a clear tag create noise. The dispatcher reviews the queue and sees a contact with no urgency flag, no service type, and no reason why it’s there. The Update CRM node removes that ambiguity: the contact is tagged, the reason is logged, and the workflow is closed. Monthly reports on disqualification reasons also tell you whether your ad targeting is generating out-of-area volume worth investigating.
What to check: Pull a CRM report filtered by “Disqualified” after a test run. Confirm the reason field is populated and the contact doesn’t appear in the active lead queue.
The complete qualification workflow
contact.created → AI Extract (service type, system status, equipment age, property type, service area, timeline) → AI Decision (qualified-high-urgency / qualified-routine / disqualified) →
Path A (high-urgency): Create Task (urgent dispatch) → Send SMS (60-second confirmation) → end
Path B (routine): Send SMS (confirmation + booking prompt) → Create Task (standard queue) → end
Path C (disqualified): Send SMS (polite close) → Update CRM (disqualified tag) → end
What the Dispatcher Sees?
The output of this workflow is a queue that’s already sorted before the dispatcher opens it.
Path A leads appear at the top with an urgent flag and a task body that includes all six qualification fields. The dispatcher sees: “AC fully down, unit is 11 years old, homeowner confirmed, in-area, needs today.” They know the job, the urgency, and the replacement conversation potential before making the call.
Path B leads appear in the standard queue with a due date, service type, and equipment age. The dispatcher works through them in order without triaging by call log timestamps.
Path C leads don’t appear in the active queue at all. They’re in the CRM, tagged and closed, available for reporting but not requiring dispatcher attention.
Operators running this workflow save over 10 hours per week previously spent on manual triage, call note review, and morning queue organization.
How ServiceAgent Is the 24/7 AI Office Manager?
ServiceAgent’s AI voice agent answers every inbound HVAC call and conducts the qualification conversation in real time, including calls at 11pm and on weekends when your office is closed. The six criteria are captured during the call itself, not after review. The Workflow Builder reads those criteria and routes the lead before the homeowner is off the call.
The qualification conversation your CSR used to run on business hours runs on every call, every day, automatically. 75% of AI-handled calls convert to confirmed appointments because the lead is captured, confirmed, and routed before the homeowner considers calling the next company.
Book a demo or sign up free and build the qualification workflow today.
Frequently Asked Questions
What questions should the AI ask to qualify an HVAC lead on the phone?
The AI voice agent should capture six criteria on every inbound call: service type (repair, replacement, maintenance, or new install), system status (fully down, degraded, or functional), equipment age, whether the caller is the homeowner or a renter, whether their service address falls within your coverage area, and their timeline (emergency, this week, or planning ahead). These six answers determine the lead’s ticket value, dispatch urgency, and routing path. An AI that only captures service type and name is collecting contact information, not qualifying a lead.
How is AI lead qualification different from an IVR press-1 system?
An IVR system (press 1 for emergency, press 2 for maintenance) captures one data point through self-selection. The homeowner picks the category that sounds closest to their situation, which is often inaccurate. An AI voice agent conducts a natural-language conversation that captures all six criteria through questions, follow-ups, and clarification. The AI can ask “how old is your system?” and understand “it was installed when we moved in about ten years ago” in a way an IVR cannot. The difference shows up in the quality of the CRM data the workflow acts on.
What happens to leads that don’t qualify (out-of-area callers or renters)?
They enter Path C of the qualification workflow, which sends a polite SMS explaining the situation (out-of-area or landlord authorization required) and tags the contact as disqualified in the CRM. This is different from routing them to voicemail, which leaves the caller with no explanation and no next step. A renter who receives a clear “this requires landlord authorization” message is more likely to loop in their landlord than one who got voicemail and assumed the company wasn’t available. Monthly disqualification reports by reason also help identify whether ad targeting is generating out-of-area volume worth adjusting.