The Home Services Call-to-Job Gap

There is no single missed-call rate

CallRail's 2025 benchmark report says 14% of home-services calls were missed. Its full dataset covered 1.1 million de-identified leads or conversations across seven industries, but it did not disclose the home-services sample or define “missed.” Invoca reports 27% among home-services companies using its tracking platform, based on “thousands” of calls without a sample size, period or sampling frame.³¹

CallRail separately says 28% of calls to businesses go unanswered, but that figure comes from a product-beta cohort and is not HVAC-specific. Workiz's 40% and Scorpion's 30–50% claims appear in product-launch marketing without usable source methods. Those latter figures are not published here as benchmarks.

The pattern is uncomfortable: the larger the claimed problem becomes, the less methodology is disclosed and the closer the claim sits to a product launch. That does not prove the lower figure is correct. It means the sources cannot support a universal rate and should not be averaged.

The commercial question is booking, not answering

An answered call is not yet a job. The caller must represent a service opportunity, the company must qualify the need, capacity must exist, the appointment must fit and the result must become a dispatchable job. ServiceTitan's platform data provides the strongest public view of this part of the funnel.

ServiceTitan reported 62% for top-performing contractors and 39% for the rest in its more recent benchmark. Its June 2022 analysis of more than 3,000 U.S. and Canadian trades businesses put the typical shop at 42%, shops with fewer than five technicians at 24%, and those with 25 or more at 59%. The 2022 trade breakout reported plumbing at 43%, electrical at 41% and HVAC at 38%.⁴⁵

These are ServiceTitan customer benchmarks, not national industry averages. Customers on a large field-service platform may be bigger and more systematized than the contractor population overall. Still, the shop-size gradient makes one point hard to ignore: booking performance is connected to staffing and operating capacity, not simply whether somebody picked up.

The after-hours finding that matters

ServiceTitan's dated 2022 analysis showed booking rates dropping sharply after 6 p.m. At larger shops, booking fell from a 61% morning peak to 21%. At smaller shops, it fell from 26% to 9%. These calls were answered; the weakness appeared later in the funnel.⁵

Possible mechanisms include lower staffing, less appointment capacity, weaker qualification, reduced decision authority or fewer scheduling options. The public evidence does not establish which mechanism dominates. It does show that after-hours performance is not only an answer-rate question.

After-hours HVAC volume is definition-sensitive

ServiceTitan analyzed aggregated, anonymized call records from participating customers across 2025. It used a tenant-weighted same-store cohort and defined after hours as everything outside 8 a.m.–5 p.m. local time, Monday through Friday. The sample size was not disclosed.⁶

Ruby reported 3.8% of calls to about 14,000 small businesses arriving after hours in 2021, from 25.3 million calls. But Ruby defined after hours as 6 p.m.–5 a.m. and covered more than 14 industries—not residential HVAC specifically. The values are not contradictory and do not form a range. One definition includes 5–6 p.m. and all weekend; the other does not.⁷

The capacity paradox

Jobber's December 2025 survey of 1,050 U.S. home-service business owners found 80% saying they were fully booked or close to it, while 3% reported trouble filling schedules. Jobber sells an AI receptionist and has a commercial interest, but it disclosed the survey provider, field date, sample and margin of error.⁸

If a contractor is already constrained by technicians, appointment capacity or service geography, the marginal value of capturing one more call is not automatically the full value of a job. A missed call may be lost incremental revenue. It may also be displaced demand, deferred demand, a low-value request or a job the company could not fulfill.

The folklore statistics

Bad statistics survive because they are commercially useful. They turn a complicated funnel into a simple loss number and make the purchase decision feel urgent. This audit traced several widely repeated claims.

The 411 Locals example illustrates mutation. Its small 2016 study said 62% of observed calls received voicemail or no response. Later content turned that into “62% will not leave voicemail,” a different claim. The five-minute rule illustrates scope inflation: evidence about outbound response to B2B web forms became advice about inbound consumer phone calls.

The AI category is real

Product launches, distribution and investment make clear that AI reception in home services is no longer hypothetical. They do not independently prove its outcomes.

What AI products claim

The sharpest juxtaposition comes from ServiceTitan itself. Its platform benchmark says top-decile human operations booked 62% of inbound lead calls. Its AI pilot page says 70% overall and 90% when capacity-adjusted. If comparable, the agent would outperform the top decile of human operations. The public sources do not disclose enough to establish comparability. The result is consequential and unproven—not evidence that AI beats humans.⁴⁹

The consumer contradiction

Housecall Pro surveyed 1,040 U.S. homeowners through SurveyMonkey Audience in October 2025 and found 53% comfortable with AI handling initial inquiries. Respondents were screened as household decision-makers and balanced to U.S. Census standards—the strongest disclosed consumer method in this review.¹⁰

ServiceForge, which sells human answering, reports roughly 85–87% preferring a real person from a OnePoll survey of 6,000 adults. Its report page and press release conflict on both field date and result: a January release describes 85%, while a report page says the survey ran in May and reports 87%. The country, screening and weighting are not stated.¹¹

The questions are also different: comfort with AI for an initial inquiry is not the same as preference for a person. Neither result cancels the other. Vendor-sponsored consumer research points toward the sponsor's commercial model, making independent research especially valuable.

Phone matters, but it is not the only channel

Modernize's homeowner research reports first-contact preferences of roughly 30% phone, 29% text, 26% email and 11% online booking. Phone preference fell from roughly 37% the prior year. The source does not publish enough method detail to support two-decimal precision, so rounded values are more responsible.¹²

The initial-contact market is fragmenting. That does not mean phone is dying: it remains the largest single channel in this dataset. It also does not mean phone is everything. Call handling increasingly sits inside a broader system of text, email, booking links and follow-up.

Labor context: field demand rises as CSR employment falls

BLS covers the national occupations, not HVAC call centers specifically. It does not say voice AI is replacing HVAC CSRs. The safe structural conclusion is that field-service labor demand is projected to grow while the broader CSR occupation is projected to shrink.¹³¹⁴

ACCA and Farmington Consulting Group survey more than 1,000 HVACR contractors and say their Contractor of the Future work includes callback and lead-follow-up benchmarks. The detailed figures were not publicly accessible in the sources reviewed, so none are inferred here. Acquiring that study is a priority for future research.¹⁵

What is actually measured

Only pieces of the call-to-job chain have public measurement. Qualification, schedule fit, dispatch coordination, customer updates, exception handling and human handoff are described extensively by vendors but rarely measured. That is the central evidence gap.

What should be measured next

1. Calls per booked job, separating qualified demand, non-service requests and abandoned interactions.
2. Booking rate by hour and day in 2026, using stable definitions and capacity data.
3. Callback behavior after missed calls, by time, channel and job type.
4. Voicemail abandonment in an independent multi-operator sample.
5. After-hours share under standardized 8–5, evening, overnight and weekend definitions.
6. Matched AI-versus-human booking with comparable call types, capacity and policies.
7. CSR time and dispatcher interruptions from observed work rather than estimates.
8. Incremental versus displaced revenue when technician and schedule capacity are constrained.

Measure the whole path to a job

Home services is automating the call-to-job workflow faster than it is standardizing how that workflow is measured. The strongest public evidence is not a dramatic universal missed-call rate. It is the measured difference in booking performance, the collapse in after-hours conversion within dated platform data, the seasonal variation in calls outside business hours and the capacity constraint reported by operators.

The industry's biggest opportunity is not another unsupported claim about how much every missed call costs. It is better measurement of booking, capacity, follow-up, coordination, handoff and incremental revenue. Until those measures exist, vendors can prove that the category is active—but not that one set of headline outcomes applies across the industry.

Methodology

Claire Research reviewed public sources through August 31, 2026, including government labor data, association material, independent trade reporting, operational-platform analysis, vendor surveys, product announcements and qualitative community evidence. The evidence register contains 58 items and preserves conflicting rows, weak methods and claims marked unsafe to publish.

Numbers were checked against the cited source. Vendor data is labeled. Platform populations were not treated as national samples. Conflicting figures were not averaged. Marketing arithmetic, untraceable attributions and scope-inflated claims were excluded as factual benchmarks. Community evidence was used only for qualitative context. The review does not claim peer review or independent outcome testing.

Download the cleaned 58-row evidence register (CSV).

Selected sources

1. Invoca, home-services missed calls — vendor platform evidence.
2. Forbes contributor article, 2014 — origin of an unsourced voicemail claim.
3. CallRail 2025 benchmark report — vendor platform evidence.
4. ServiceTitan top-performer profile — platform data through January 2026.
5. ServiceTitan call-booking analysis — June 2022 platform data.
6. ServiceTitan HVAC after-hours analysis — 2025 platform cohort.
7. Ruby call-volume report — answering-service platform data.
8. Jobber Home Service Trends Report — vendor-sponsored survey, 1,050 owners.
9. ServiceTitan AI Virtual Agent — vendor pilot claims.
10. Housecall Pro homeowner survey — vendor-sponsored survey, 1,040 homeowners.
11. ServiceForge AI survey report — commissioned survey with date/number inconsistency.
12. Modernize homeowner communication preferences — vendor research.
13. U.S. BLS, HVACR mechanics and installers — government statistics.
14. U.S. BLS, customer service representatives — government statistics.
15. ACCA Contractor of the Future study — association research; detailed metrics not publicly accessible.