Research / HVAC intake
Capture is only
the first step.
Across 10 Florida HVAC businesses, public contact paths were already common. The pattern worth examining was what comes next: qualification, scheduling and the handoff to a person.
Research by Automataria ·
What we looked at, and how.
This directional study reviewed a convenience sample of 10 Florida HVAC businesses on or around September 8, 2026. The sample was not random and is not statistically representative of the HVAC industry.
The sources were public company websites, booking and contact flows, published office and service hours, a ServiceTitan customer page and one dispatcher job listing. No interviews were conducted. No CRM, dispatch board, phone system, inbox, private analytics or internal workflow was accessed.
Contact options were already present.
Every business in the sample exposed a public phone path. Most also exposed another way to make contact: a web form, text, WhatsApp, callback or online-scheduling path. This suggests investigating the step after capture rather than assuming capture is absent.
For example, Cool Seasons exposed service requests and call-or-text options; MR A/C OF ORLANDO combined scheduling prompts with an emergency line; Air Florida Solutions offered phone, WhatsApp and a form. These are observations about published entry points, not measurements of how requests were handled.
A scheduling prompt can still lead to a handoff.
English Air’s scheduling form stated that the company would contact the customer within one business day. In this example, a service-scheduling submission led to a stated human follow-up, not necessarily an immediate confirmed appointment.
Office availability and service availability could also differ. Advanced Air separated normal business hours from after-hours coverage; 3H AC published weekday office hours alongside weekend service; Dynamic listed business hours while advertising 24/7 service. These boundaries raise questions about handoffs. They do not establish delays or missed requests.
Different tools coexist with human coordination.
The visible software signals were mixed. Dynamic identified its site as powered by OneLocal; ServiceTitan featured The A/C Guy of Tampa Bay on its SMB software page. Other businesses exposed different or undisclosed tooling. This evidence does not support tying the intake problem to one CRM.
Public descriptions also showed human coordination roles: Bob Heinmiller’s dispatcher listing covered scheduling and coordinating service calls and requested ServiceTitan experience. Anthony’s team page listed operations, CSR management, customer-service and systems roles. These sources describe roles alongside software; they do not reveal the complete internal process.
What these observations cannot tell us.
This is a snapshot of public intake surfaces. Websites, team pages and job listings can change after the observation date. Published hours or follow-up promises are not evidence of actual response times, staffing at every hour or completed bookings.
The sample cannot establish conversion rates, lost revenue, market share or the quality of any company’s internal operation. It does not represent the industry statistically. The useful result is a set of directional patterns and questions to test, not a ranking of the businesses.
The product hypothesis: test the step after intake.
Our inference from these observations is to test the path Inbound Lead → Qualified → Booked Service, rather than assume that the main problem is a missing CRM or lead-capture channel.
The first narrow capability to validate is to normalize an incoming request, classify service type, urgency and service-area fit, identify missing information, then recommend the next action. Missed-call recovery is one possible entry into that same workflow. This is Automataria’s product hypothesis, not an observed result of the study.
Shadow mode is the validation method, not a proven outcome.
The next test would observe a copy of incoming leads without taking customer-facing action, produce classifications and a recommended next step, then compare them with the human CSR or dispatcher decision. External actions would be enabled only after evidence supports doing so.
This study does not show that Automataria has improved accuracy, conversion, response times or bookings. Those claims would require later customer or pilot evidence.
Where does the handoff get stuck?
Tell us how requests arrive, who handles them and what should happen next.
Tell us about your intake