Skip to main content
zatersio

AI Receptionist for Trades: What It Does and What It Costs

AI Receptionist for Trades: What It Does and What It Costs

Sketch of phone and trade tools

An AI receptionist for trades answers your phone when you can’t, collects the job details, and sends them to you or your calendar. It works while you’re under a house, up a ladder or asleep. It won’t replace your judgment on a tricky quote or an angry customer. It handles the predictable first step so nobody hits voicemail and hangs up.

This guide covers what the tool does day to day, where it fails, the privacy rules that apply in Australia, what a build costs, and how to decide between an off-the-shelf product and a custom one.

Zatersio
Turn Missed Calls Into Captured Enquiries
Zatersio builds bespoke AI-driven automation for trades, helping capture customer enquiries and reduce repetitive administrative work.
Explore Zatersio

TL;DR

  • An AI receptionist picks up every call, captures name, address, job type and urgency, and books or flags the job. Humans stay in charge of quotes, disputes and emergencies.
  • Design it so complex or upset callers transfer to a person. Don’t treat it as set-and-forget.
  • If it records or processes voice, the Privacy Act 1988 applies, so you need a clear privacy policy and sensible handling of the data.
  • Our AI Phone & Enquiry Agent is a fixed-price one-off build, from $2,000 to $15,000 AUD depending on scope, built by engineers in Melbourne.
  • Start with one job: answer, capture, notify. Add booking and quoting later.

Table of Contents

Why missed calls cost trades more than most businesses

A plumber with a burst pipe call doesn’t leave a voicemail. They ring the next number. Trades work is urgent, local and often decided in the first call, and the person who could answer is usually holding a tool.

That’s the gap an AI receptionist fills. It isn’t about looking fancy. It’s about the three minutes between a customer dialling and a competitor picking up.

Some numbers worth knowing, with caveats. Surveys of small businesses using AI in customer service report gains in after-hours availability and faster resolution times, though the same source stresses that people still do better on personal, complex and emotional conversations. Treat the percentages as indicative, not a promise for your business. Your own call log is the better evidence: count how many calls you missed last month and what an average job is worth.

A quick sum, using your numbers, not ours:

  1. Missed calls per month that were new enquiries (check your phone log).
  2. Share you’d normally convert into jobs (say, one in three).
  3. Average job value.

Multiply them. If the result is a few thousand dollars a month, a one-off build pays back quickly. If you miss two calls a month, you probably don’t need one yet.

What an AI receptionist actually does for a tradie

A good one does a short list of things well:

  • Answers instantly, day or night, in a natural voice, with your business name.
  • Captures the essentials: caller name, number, suburb or address, what’s wrong, how urgent.
  • Sorts the call: emergency, quote request, existing customer, supplier, sales spam.
  • Sends you a summary by text or email, or drops the job into your calendar or job system.
  • Handles the routine questions: service area, hours, whether you do a particular job.
  • Transfers or flags anything it shouldn’t handle.

Note what’s missing: pricing negotiation, diagnosing a fault over the phone, settling a complaint. Those stay with you.

The “predictable first step” rule

Research from Wharton makes the point neatly: experts advise giving the AI the predictable first step and leaving the judgment call to people. Confirming hours or collecting details suits automation. Billing disputes and irate customers don’t.

For a trades business that translates to simple rules:

  • Gas smell, water through the ceiling, no power: ring the on-call person straight away.
  • Caller says “I’m not happy with the last job”: transfer or flag for a same-day callback.
  • Anything the agent isn’t sure about: take the details and promise a callback, never guess.

Where it goes wrong

Being honest here saves you money. An AI receptionist is a poor fit when:

  • Your callers are mostly upset or in distress. Some will want a human straight away. Offer a clear way to reach one.
  • Nobody follows up. If the summaries land in an inbox nobody reads, you’ve bought a faster way to lose jobs.
  • The script is thin. A bot that doesn’t know your service area or the jobs you refuse will frustrate callers.
  • It’s treated as set-and-forget. A Talkdesk survey of US small businesses found over 40% prioritise human oversight, usage guardrails and staff training. The lesson carries over: check call summaries weekly at first and tune the agent.

Some callers dislike talking to a bot, and that’s fair. The fix is to be upfront that it’s an AI assistant, keep it brief, and make the handover to a person easy.

Privacy: what Australian trades need to get right

If your agent hears a caller’s name, address and phone number, you’re collecting personal information. The Privacy Act 1988 and the Australian Privacy Principles apply to AI tools that transcribe or process voice, according to one industry summary. Check the details against official guidance, and note that the Act may not cover every very small business, so confirm your own position with a privacy adviser or the OAIC.

The practical points:

  • Say it’s an AI at the start of the call and that details are recorded or processed.
  • Keep your privacy policy current and clear about how AI tools collect, hold and use data.
  • Avoid voice verification. The OAIC classes biometric information, including voice patterns used to verify identity, as sensitive information that needs explicit consent and higher protection.
  • Know where the data sits. Some products send recordings to overseas servers. Ask the question before you sign.

On that last point, we build on Australian-resident or client-owned infrastructure, and you choose data residency. That makes the answer to “where does my customers’ voice data go?” a short one. Recording laws also vary by state and territory, so check the rules where you operate before you record calls.

Off-the-shelf, human service or custom build?

There are three broad ways to get calls answered.

Option Strengths Trade-offs
Live answering service staffed by people Warm human voice, handles odd calls Per-call or per-minute fees, scripts are generic, limited to what the operator knows
Subscription AI answering product Quick to switch on, low entry cost Generic flows, monthly fees that never end, limited say over data and integrations
Custom-built AI agent Fits your jobs, your tools, your tone; you own it Higher upfront cost, needs a proper scope

For pricing context: an in-house receptionist is a very different spend. One Australian provider puts the all-in cost at roughly $71,000 to $85,000 a year once salary, super and leave are counted. A full-time hire makes sense once you have an office, volume and admin to match. For a crew of three to ten on the road, it rarely does.

The broader industry direction is worth noting. Phone tools such as Zoom’s phone product now include always-on AI receptionists, and the push is toward agents that connect to your CRM, scheduling and billing, so context carries from a text enquiry to a phone call. That’s the case for a build that plugs into your stack, rather than a stand-alone answering box.

What a build costs with us

What a build costs with us — overview diagram

We publish our pricing as fixed one-off ranges, not per-call meters. For this keyword the relevant offer is the AI Phone & Enquiry Agent, at $2,000 to $15,000 AUD one-off. Where you land depends on how much the agent has to do.

Our wider pricing page sets out the related builds:

Build Price (AUD, one-off)
Single workflow automation $2,000 to $4,000
Multi-step automation across your stack $4,000 to $9,000
Custom AI agent build $9,000 to $15,000
AI Phone & Enquiry Agent $2,000 to $15,000

What moves the price within the range:

  • Answer and notify only. Pick up, capture details, send a summary. This is the simple end.
  • Booking into a calendar or job system. More moving parts, more testing.
  • Several channels. Phone plus web chat or text enquiries, with context kept across them.
  • Custom rules. Emergency routing, after-hours on-call lists, service-area checks.

Optional ongoing maintenance is available if you want us to keep tuning it. We don’t publish a price for that, so ask when you scope.

How we’d build it

A typical project runs like this:

  1. Scope call. We ask what you do, where you work, which jobs you want and which you refuse, and what happens to an emergency.
  2. Fixed price. You get a number once we understand the job, not a meter that runs.
  3. Build. Engineers in Melbourne write it. No account manager in the middle, no offshore handoff.
  4. Test with real calls. You ring it, we tune the script, the routing and the edge cases.
  5. Go live and review. Check the first weeks of summaries together and fix what the callers actually said.

You own the code and IP, and it runs on Australian-resident or client-owned infrastructure.

A note on the R&D Tax Incentive

People ask whether an AI receptionist qualifies for the R&D Tax Incentive. Usually, a standard one doesn’t. Government guidance on AI-related activities explains that eligibility depends on the work itself, not on the label “AI”. The software development sector guide describes a core R&D activity as one where you hit a specific technical hurdle that existing knowledge can’t resolve, even for experienced professionals.

In plain terms: wiring an off-the-shelf voice tool to your calendar is routine development. A bespoke agent that needs real experimentation to work is more likely to be relevant. Speak to your accountant or an R&D adviser before you count on it. We can structure eligible software projects, but we won’t tell you a simple answering flow qualifies when it doesn’t.

Routine integration versus bespoke agent experimentation

Our view: start small, keep a human in the loop

We’ve seen the same pattern with trades and service businesses. The owners who get the most from automation don’t begin with a grand agent. They begin with the single most annoying leak in their day: the missed call, the re-typed job card, the quote that waits a week.

Our advice:

  • Fix the first thirty seconds. If every caller reaches a clear, friendly voice that captures their details, you’ve already won most of the benefit.
  • Write down your rules before we write code. Which calls are emergencies? Which suburbs do you service? What do you never quote over the phone? The agent is only as good as that list.
  • Keep people for judgment. Quotes, complaints and anything emotional go to you or your office.
  • Own what you build. Subscriptions are fine for a trial. If this becomes the front door to your business, you want the code, the data and the choice of where it lives.

If the numbers say you’re losing jobs to voicemail, don’t wait for the perfect plan. Build the simple version, review it in a few weeks, and extend it.

— Lakitha

Make the after-hours calls count

Answering the call is half of it. If you also pay for ads to make the phone ring, you want those calls caught, whatever the hour. Pairing a call-capturing agent with well-run paid search makes sense, and an agency such as Senior Ad Managers’ Google Ads service can help tie campaign spend to the revenue that actually comes in. Do the call-handling fix first, so you aren’t paying for clicks that end in voicemail.

Ready to stop losing calls?

If you run a trade business and want an agent that answers, captures and routes your calls, look at our AI Phone & Enquiry Agent. Tell us what you do and how your calls go today, and we’ll scope a fixed price. If you want to compare it with other builds, see our pricing or read how AI agents compare with hiring a VA.

FAQ

Will an AI receptionist replace a human receptionist?

For most trades it replaces the missed call, not the person. Research on small businesses points to AI complementing human teams by handling routine interactions, while people take the high-value and emotional work. If you have an office team, the agent frees them from first-touch calls.

How much does an AI receptionist cost in Australia?

It varies by provider and call volume, from low monthly subscriptions to packages in the thousands. With us, the AI Phone & Enquiry Agent is a one-off build from $2,000 to $15,000 AUD, depending on how much it needs to do. Optional maintenance is separate.

Do I have to tell callers it’s an AI?

You should be upfront. The Privacy Act applies when AI tools collect personal information, and a clear privacy policy and notice about how calls are handled are part of doing that properly. Check recording rules in your state or territory too, and ask a privacy adviser if you’re unsure.

Does an AI receptionist qualify for the R&D Tax Incentive?

Not automatically. Eligibility depends on a genuine technical hurdle and systematic experimentation, and routine integration of off-the-shelf tools is generally excluded. Bespoke agent work is more likely to qualify, so confirm with your accountant or an R&D adviser.

Sources