Under 5 trucks, missed calls are the whole problem: an AI answering service. Fastest to switch on, lowest month-one cost, and call handling is genuinely the thing it does well.
Already on Jobber, Housecall Pro, or ServiceTitan and the gaps are small: turn on what your FSM already includes before buying anything. Most contractors are paying for automation features they have never enabled.
5 to 25 trucks with three or more workflows leaking money: a fixed-price build. The economics turn over around the 18-month mark against any subscription-based option.
You have an ops person with real time and appetite: DIY on Make or n8n. Cheapest in dollars, most expensive in attention, and the maintenance never ends.
Skip retainer agencies unless you genuinely need continuous change. Most HVAC automation is a build-once problem, and a retainer prices it as an ongoing one.
What HVAC contractors actually automate
Before comparing providers, get specific about the job. Across HVAC builds the same five workflows come up, and they are worth different amounts:
- Missed-call text-back. The single highest-value automation in the trade. A missed call outside hours is a lead that calls the next contractor. An automatic SMS inside 60 seconds keeps it alive.
- Maintenance agreement renewals. Recurring revenue that quietly decays because nobody has time to chase it. Renewal reminders on a schedule recover a meaningful share.
- Review requests timed to job completion. Asking two days after the tech leaves converts far better than asking whenever someone remembers.
- Invoice follow-up. Unpaid invoices age because chasing them is nobody's favorite job. A sequence does it without the awkwardness.
- Estimate follow-up. Estimates that go out and never get a second touch are the most common leak in a growing shop.
See the full missed-call text-back workflow for how one of these is actually built, including the failure modes.
The five provider types
| Provider type | Typical cost | Time to live | Who owns it | Best fit |
|---|---|---|---|---|
| AI answering service | Monthly subscription, often per-minute or per-call | Days | Vendor | Call handling only, under 5 trucks |
| FSM built-in AI (Jobber, Housecall Pro, ServiceTitan) | Included in your existing plan, or a paid tier up | Hours | Vendor | Anyone already paying for the FSM |
| Retainer automation agency | Monthly retainer, ongoing | Weeks | Usually the agency | Continuous change, larger operations |
| Fixed-price build shop | One-time project fee | 1–3 weeks | You | 3+ workflows, 5–25 trucks |
| DIY (Zapier, Make, n8n) | Platform fees plus your time | Weeks to months | You | In-house ops capacity exists |
1. AI answering services
These pick up the phone when you cannot. Modern ones book appointments, qualify the caller, and hand off to a human when the conversation goes sideways.
What they do well: the narrow job of not losing a call. Setup is genuinely fast, and for a small shop where the owner is on a roof half the day, that alone can pay for itself.
Where it breaks: the call is only the first step. The service answers, but your maintenance renewals, invoice chasing, and review timing are untouched. Contractors often end up stacking three subscriptions to cover what one build would have handled, and the per-minute pricing scales against you in your busy season, which is exactly when you can least afford surprises.
Wrong choice when: your problem is not actually the phone.
2. Your FSM's built-in AI
Jobber, Housecall Pro, and ServiceTitan all ship automation features. Jobber's plans run $49/mo Core, $139/mo Connect, $199/mo Grow, and $499/mo Plus on monthly billing. Housecall Pro runs $79/mo Basic, $189/mo Essentials, and $329/mo MAX on monthly billing, with lower annual rates.
What they do well: it is already in your stack, already has your customer data, and costs nothing extra if you are on the right tier. This is the first thing to check and the most commonly skipped.
Where it breaks: it stops at the edge of the platform. The moment a workflow needs to touch a tool the FSM does not integrate with, you are out of options. Feature depth also varies sharply by tier, and the automation you want is frequently two tiers up.
Wrong choice when: your workflows span tools. See the HVAC CRM comparison for what each platform covers.
3. Retainer automation agencies
A monthly fee for ongoing automation work, usually with a set number of hours or workflows.
What they do well: genuine value when your operation changes constantly and you need someone on call to change with it. Larger multi-location operations often do need this.
Where it breaks: most HVAC automation is a build-once problem. Missed-call text-back does not need monthly iteration; it needs to work. A retainer converts a one-time cost into a permanent one, and the automations frequently live in the agency's accounts, which means leaving means rebuilding.
Wrong choice when: your processes are stable, which for most contractors they are.
4. Fixed-price build shops
A scoped project, a fixed number, and a handoff. This is the category Aplos AI is in, so treat the following with appropriate suspicion and check it against the others.
What it does well: the cost stops. You own the workflow and it runs on your accounts, so there is no monthly fee to the builder and no hostage situation. For three or more workflows it is usually the cheapest option by year two.
Where it breaks: the upfront number is real money at a point when you have not seen results yet, and that is a legitimate reason to start smaller. Scoping matters more than it does with a subscription, because a badly scoped build is an expensive mistake rather than a cancelled plan. It is also the wrong shape if you genuinely need someone iterating with you every month.
Wrong choice when: you have one narrow problem that an off-the-shelf tool already solves. If missed calls are your only issue and your FSM handles it, buy nothing.
5. DIY on Zapier, Make, or n8n
What it does well: lowest cash cost, complete control, and you learn your own systems properly.
Where it breaks: the build is the small part. Budget 40 to 60 hours to design, build, test, and maintain one non-trivial workflow, and the maintenance share recurs forever. Task-based pricing also compounds quietly at volume.
Wrong choice when: the person who would build it is the same person running the business.
What this costs by year three
The month-one comparison flatters subscriptions and the three-year comparison flatters builds. Run the numbers on the horizon you actually plan to be in business for:
| Option | Year 1 | Year 3 cumulative | Asset at end |
|---|---|---|---|
| Three stacked subscriptions | Lowest | Highest | Nothing |
| Retainer agency | Moderate | Highest | Usually nothing |
| Fixed-price build ($8,000–$25,000 for a multi-tool build) | Highest | Lowest | The workflows, on your accounts |
| DIY | Lowest in cash | Low in cash, high in hours | The workflows, plus the maintenance |
Our full four-way breakdown with the underlying assumptions is on the comparison page.
Questions to ask any provider
- Whose accounts does this run on? If the answer is theirs, you are renting, whatever the pricing page says.
- What happens when we stop paying? The honest answer separates the categories faster than anything else.
- Show me the exception path. Anyone can demo the happy path. Ask what happens when the API is down or the customer replies something unexpected.
- Who fixes it when it breaks at 7pm in August? Peak season is when automation failures are most expensive.
- What is the total at 36 months? Not the monthly. The total.
Choosing by company size
- 1–4 trucks: enable your FSM's built-in automation first. Add an answering service if calls are still leaking. Buy nothing else yet.
- 5–15 trucks: the crossover zone. If three or more workflows leak money, a fixed-price build is usually cheapest by month 18.
- 15–25 trucks: a build, and budget for integration work between systems that were never designed to talk.
- 25+ trucks or multi-location: this is where a retainer or an in-house hire starts to genuinely make sense.
Not sure which category fits your shop?
We run a free 30-minute audit that maps your current workflows and tells you what is automatable. If the answer is "turn on the feature you are already paying Jobber for", we will tell you that and you will not hear from us again.
Get a Free Automation Audit →