AI Voice Agent for Home Services: Dispatch, Emergency Triage, Recall

By Oskar Korjus, co-founder of luup voice agents for home services. Numbers come from luup's audit of 11 mid-market home services contractors and patterns from our 47-deployment voice agent audit (where home services ranked second-worst at 80% failure rate, mostly from emergency triage gaps and dispatch integration drift). Industry baseline from ACCA HVAC industry data and ServiceTitan industry research.

1. Where home services contractors leak inbound

Run the math on your peak week. Across the 11 home services contractor audits luup ran in 2026, median peak-season voicemail rate was 38%. Median callback time on missed calls was 2-4 hours - by then the customer has called a competitor. Worst leak windows: weekday 7-9am (morning emergency rush), weekday 5-7pm (homeowners discovering issues after work), Saturday 8am-12pm (weekend project discoveries), after-hours emergencies (2-6am).

The leak compounds across three structural failures. First, dispatch capacity is fixed (you have N trucks); inbound volume varies 3-5x peak vs off-peak. Second, the emergency-vs-scheduled triage requires trade-specific judgment that intake staff often lack - resulting in either non-emergencies dispatched as emergencies (wasting truck capacity) or true emergencies queued as scheduled work (creating customer harm and liability). Third, preventive maintenance contracts are massively under-utilised; most contractors have 200-800 active maintenance contracts but only convert 25-35% of seasonal recalls because the front office cannot make outbound calls during peak.

Home services ranked second-worst in the cross-vertical 47-deployment voice audit at 80% failure rate, mostly from emergency triage gaps and dispatch integration drift. The fix patterns from that audit apply directly: documented triage flows with named on-call humans, real-time dispatch system integration, weekly transcript review for trade-specific vocabulary tuning.

2. Five loops a home services voice agent closes

Five loops cover the operational surface. Each has defined triggers, integration paths, success metrics.

2.1 Loop 1 - Inbound capture (peak + after-hours)

Trigger: any call to the main line that goes unanswered, or any call after business hours. Data path: Twilio inbound to Vapi or Retell agent to ServiceTitan, Housecall Pro, or Jobber for job creation. Success metric: 95%+ inbound calls answered within 6 seconds.

2.2 Loop 2 - Emergency triage and dispatch

Trigger: caller mentions safety risk (gas leak, water damage, electrical hazard, fire). Data path: 4-question triage to severity score to immediate dispatch with 15-minute SLA plus on-call manager alert. Success metric: 100% true emergencies dispatched within window.

2.3 Loop 3 - Scheduled work booking

Trigger: caller is non-emergency and accepts scheduling. Data path: dispatch system schedule API to slot held to confirmation SMS. Success metric: booking completed within 90 seconds.

2.4 Loop 4 - Preventive maintenance recall

Trigger: dispatch system scan for maintenance contract due dates plus seasonal triggers. Data path: outbound call queue with seasonal tune-up offer. Success metric: 35-55% conversion on active maintenance contract holders.

2.5 Loop 5 - Post-service follow-up and review

Trigger: 24-48 hours after job completion. Data path: outbound call for satisfaction confirmation plus review request plus payment status. Success metric: 35-50% review collection rate, payment-on-time rate above 90%.

The Loop Map Generator walks an operator through scoping all five for a specific contractor in 8-10 minutes.

3. Deep dive: emergency triage end to end

Emergency triage is the loop with the highest patient-safety stakes. Across the 11 contractor audits, we documented 5 cases where a true emergency caller (gas leak, active flooding, exposed live wires) was queued as a scheduled callback rather than dispatched immediately - one led to property damage liability, two led to customer churn to competitors who responded faster.

The 4-question triage script:

1. "Tell me what is happening - what are you noticing?" Open-ended question to surface the issue and any embedded distress signals.
2. "Are you noticing any of these: smell of gas, active water damage, sparks or smoke, anyone in the home with breathing difficulty?" Yes to any auto-dispatches with 15-minute SLA.
3. "How long has this been happening?" Combined with severity to determine urgency tier.
4. "Are you safe to stay in the home, or should you wait outside?" Final safety check; gas leaks and major electrical issues route to "leave and call us back from outside" guidance.

Routing logic. Gas leak signals (smell, hissing, headaches): immediate dispatch plus instruction to evacuate plus call gas company. Active flooding: immediate dispatch plus instruction to shut off main water valve. Electrical hazard (sparks, smoke from outlet, exposed wire): immediate dispatch plus instruction to shut off breaker. No-heat in winter (under 32F outside, household with elderly or infants): same-day priority within 4 hours. No-cooling in summer (over 90F, household with vulnerable members): same-day priority within 4 hours. Routine: scheduled within 24-48 hours.

4. Deep dive: dispatch routing end to end

Dispatch routing is where most home services voice deployments fail or fly. Manual dispatch by a front-desk operator misses three optimisations: closest-truck routing (which tech is geographically nearest), skill-match routing (HVAC tech for HVAC call, plumber for plumbing), and capacity-balance routing (load-leveling across trucks rather than overloading the favourite tech).

The voice agent integrates with the dispatch system to access real-time truck location and availability. ServiceTitan's API exposes truck GPS plus current job plus skill profile. Housecall Pro exposes similar via their REST API. The agent passes the new job request through three filters: closest-truck (within 30-minute drive of the caller's address), skill-match (the trade required matches the tech's certification), capacity (truck is not already at capacity for the day).

Three outcomes:

• Match found. Agent confirms slot with caller, dispatches the job to the tech's mobile dispatch app, sends confirmation SMS to caller with tech name plus ETA window. ~75% of calls land here.
• No same-day match. Agent offers next available slot (typically 24-48 hours out for routine work, same-day for urgent), confirms with caller. ~22% of calls.
• Specialty work outside trade scope. Some calls require specialty work the contractor does not handle (commercial-grade boiler, septic, well pump). Agent professionally explains and offers referral resources where allowed. ~3% of calls.

The dispatch optimisation alone improves truck utilisation 12-18% in our test data because closest-truck routing reduces drive time and skill-match routing reduces "wrong tech sent" callbacks.

5. Dispatch system integrations

Six platforms cover 90%+ of mid-market home services. The choice typically follows what the contractor already runs.

5.1 ServiceTitan

Dominant in mid-market HVAC, plumbing, electrical. Deepest API for dispatch, customer history, pricing, payments. Voice integration ships in 2-3 days. Premium pricing tier reflects deeper feature set.

5.2 Housecall Pro

Strong on smaller crews and trades-flexible contractors. REST API supports booking, dispatch, GPS tracking, payments. Integration takes 2-3 days.

5.3 Jobber

Best fit for sub-15-truck operations and trades-flexible smaller contractors. Simpler API, fewer features. Integration takes 1-2 days.

5.4 FieldEdge

HVAC-specific dispatch and customer management via the FieldEdge platform. Integration takes 3-4 days. Best for HVAC-only contractors with maintenance contract focus.

5.5 Service Fusion

Multi-trade contractors and commercial-residential mix. Integration takes 3-5 days.

5.6 WorkWave

Pest control, lawn care, lawn maintenance. Different operational model than HVAC/plumbing/electrical. Integration takes 4-6 days.

6. The 5-day deploy

Day 1. Dispatch system integration. ServiceTitan: credentials issued same day. Housecall Pro: developer access requested (1-2 day lag).

Day 2. Trade-specific vocabulary trained. Agent learns HVAC, plumbing, electrical terms. Triage script localised. Pricing ranges loaded for top 10 jobs.

Day 3. Emergency triage script reviewed by master tech. Recall sequence designed for seasonal triggers.

Day 4. Live integration test with synthetic calls. Recording disclosure script verified per state law.

Day 5. Live with first 10 callers. Dispatcher listens in for 2 hours. Adjustments in real-time.

7. Cost + ROI math at three contractor sizes

Profile	Inbound calls/week peak	Annual leak recovered	Monthly cost	Payback
1-2 truck owner-operator	150-350	$45-75k	€1,500-2,200	3-5 weeks
3-6 truck mid-market	400-900	$120-220k	€2,200-3,500	2-3 weeks
7-15 truck larger contractor	900-1,800	$280-480k	€3,500-5,500	1-2 weeks

Recovery scales super-linearly because larger contractors have more parallel inbound during peak windows and more maintenance contracts to recall against. Run the Revenue Leak Heatmap for your specific number.

8. Preventive maintenance recall playbook

Maintenance contract recall is the highest-margin loop because the relationship already exists. Three sub-sequences:

HVAC seasonal tune-up. Spring AC tune-up call fires April 1-15; fall furnace tune-up call fires September 15-30. Tone: warm, brief, "your AC tune-up is due; we have Tuesday or Thursday next week." Average book rate: 48-62% on first attempt for active contracts.

Plumbing annual inspection. Fires on contract anniversary. Tone: same warm-brief approach. Average book rate: 38-52% for active contracts.

Generator and standby system check. Fires monthly per contract. Higher conversion (75-85%) because customers want assurance their backup works.

Compliance note: outbound calls fall under TCPA in US. Active maintenance contract holders count as existing business relationship; cold outbound to non-customers requires explicit prior consent.

9. Five things that break home services voice deployments

1. Generic triage script. Trade-specific safety questions are non-negotiable. HVAC needs gas-leak detection; plumbing needs flooding detection; electrical needs hazard detection.
2. Manual dispatch routing. Closest-truck plus skill-match plus capacity routing is dramatically better than first-available routing.
3. Skipping the maintenance recall. Operators ship inbound capture and skip recall, missing the larger ROI wedge.
4. Pricing transparency too aggressive. Quoting fixed numbers for jobs requiring inspection breaks trust on arrival.
5. State-law gaps on recording disclosure. Some states require two-party consent. Document the disclosure script per jurisdiction.

10. Companion services for home services contractors

The voice agent closes the inbound surface. Three companion services close the operational and growth surface:

• Local services automation. The professional services automation pillar covers customer journey, follow-up, review collection, payment processing.
• Home services website on a 7-day sprint. The local services website generation pillar ships sites optimised for local search.
• Local SEO. The luup SEO pillar covers Google Business Profile, local pack rankings, review velocity.

Sibling voice-agent verticals: restaurants, dental, medspa, legal, automotive.

11. What to ship this week

Pull last week's call log. Count voicemails during morning rush and after-hours. Multiply by your average new-customer LTV. That number is your weekly leak. Run the Revenue Leak Heatmap or book a 30-minute review.

12. Frequently asked questions

13. Field notes from 11 home services engagements

Five patterns surface across the 11 home services contractors luup audited. They track structural specifics of trades operations - the seasonality, the truck-as-capacity-unit dynamic, the emergency-response stakes.

Note 1 - the dispatcher is the gatekeeper. 9 of 11 contractors had a senior dispatcher who feared the agent would replace their job. The deployments that succeeded reframed the agent as augmentation - the agent handles inbound capture and dispatch routing; the dispatcher handles tech support, customer escalations, and capacity planning. Bring the dispatcher in on Day 1 of deployment, not after.

Note 2 - peak season is asymmetric. HVAC contractors get 70-80% of annual inbound in May-September (cooling) and November-February (heating). Plumbing contractors face winter freeze events as 3-week spikes. Pre-season load testing at 5x normal volume is the only insurance against peak failures.

Note 3 - maintenance contracts are dramatically under-utilised. 8 of 11 contractors had 200-800 active maintenance contracts but converted under 30% of seasonal recalls. The agent converted 35-55% on existing contracts because outbound capacity stopped being the bottleneck.

Note 4 - emergency dispatch is the customer-loyalty surface. Customers who get dispatched within 15 minutes during a true emergency (burst pipe, gas leak) become evangelical lifetime customers. Customers who get queued during emergencies churn to competitors permanently. The 15-minute SLA pays for itself many times over in retention.

Note 5 - the after-hours service-call fee is a margin protector. Contractors who charged a documented after-hours premium ($120-280 vs $89 standard) maintained margin during off-hours dispatch. Those who did not lost money on after-hours emergency response.

The fix in every case: include dispatcher in deployment, pre-season load test, exploit existing maintenance contracts, hit the 15-minute emergency SLA, document after-hours premium pricing. Cross-vertical patterns from the 47-deployment audit generalise. Run on your specific contractor at luup voice agents for home services or book a review.

Last updated: 4 May 2026.