Hiring a VA, building an AI system, or sticking with the status quo - the honest 12-month numbers for an Australian trades or allied health business. Costs, time saved, and the trade-offs nobody walks you through.
Every Australian tradie and practice owner I audit is sitting on the same decision: the admin is eating their margin and they have to do something about it. Most pick one of three paths.
- Hire a virtual assistant. Offload the admin to a person, in Australia or offshore.
- Build an AI automation system. Run the repeatable admin on tools you already pay for.
- Do nothing. Keep absorbing the cost, hope it gets better when things slow down.
Each path has real numbers. None of them is free. Here’s what each actually costs an Aussie trades or allied health business over a 12-month window - in dollars, in hours, and in trade-offs nobody mentions when they’re selling you the option.
The numbers up front
The table below is conservative. Every figure is hedged toward the low end of what you’d actually spend, because I’d rather under-promise the saving than oversell the maths.
| Path | Year-1 cost (AUD) | Hours/week of admin still on you | Hours/week of admin still on your team |
|---|---|---|---|
| Do nothing | $0 (cash) + lost revenue from slow follow-ups | 5-15 | 10-25 |
| Hire an offshore VA (15 hr/week) | $5,000-$10,000 | 3-8 (managing the VA) | 5-10 |
| Hire an AU-based VA (15 hr/week) | $20,000-$35,000 | 2-5 | 5-10 |
| Build an AI automation system | One-off build + $40-$120/mo tools | 1-3 | 1-5 |
Cash is only one column. The other two are time. The honest answer is that most operators end up with both a VA and a system - but the order you buy them in changes the maths by tens of thousands of dollars.
Path 1: Doing nothing
The default option. Costs zero up front. Doesn’t fix anything either.
What it really costs
The cash cost is $0. The hidden cost is in revenue you never see hit the bank.
A typical Australian plumbing or electrical business waits 24-72 hours to send a quote after the initial enquiry. Industry data on home-services lead response is consistent on one point: the first business to reply wins disproportionately. In trades specifically, internal benchmarks from CRM tools like ServiceM8 and Tradify suggest the first responder closes somewhere between 40 and 60 percent of competitive quotes - even when their price is higher than the slower-responding competitor.
For an allied health practice, “doing nothing” usually means continuing to absorb a 20-30 percent no-show rate on new-patient appointments. Cliniko, Power Diary, and Halaxy all ship with native SMS reminders and recall sequences that drop that to under 10 percent. Most practices we audit have those features turned off.
What you walk away with
- Cash saved this month: real
- Cash lost over 12 months from slow quoting, missed follow-ups, and no-shows: usually 5-15 percent of total revenue
Who this path is right for
- Businesses genuinely on the way out (closing, selling, retiring inside 12 months)
- Operators with so much demand they can afford to lose 10-20 percent of leads and still hit budget
- Nobody else
Path 2: Hire a virtual assistant
The most common move. A VA is a real person who handles the admin you don’t want to do.
There are two flavours: offshore (usually the Philippines) and Australian-based.
Offshore VA: realistic 12-month cost
Typical rates for a competent Philippines-based VA in 2026 sit at $7-$12 AUD per hour, depending on agency vs direct hire and skill level. At 15 hours per week:
- Hourly cost: $7-$12
- Annual hours: 15 × 50 = 750 hours (allowing for 2 weeks downtime)
- Year-1 cash cost: $5,250-$9,000
Add a small buffer for tools the VA needs (a password manager, maybe Loom for screen recordings, a shared inbox tool) and you’re at $5,000-$10,000 all-in.
AU-based VA: realistic 12-month cost
An Australian-based VA - either a sole-trader specialist or an agency hire - runs $35-$50 per hour depending on experience and contract type. At the same 15 hours per week:
- Hourly cost: $35-$50
- Annual hours: 750
- Year-1 cash cost: $26,250-$37,500
Australian VAs are worth the premium for work that needs real judgement, local context, or live phone handling. They’re not worth it for repetitive admin a system can run.
What VAs are great at
- Complex client emails that need a human reply
- One-off compliance paperwork or NDIS submissions
- Real-time phone calls during business hours
- Anything that changes every time and can’t be templated
What VAs are bad at
- Running the same sequence the same way 50 times in a row
- Never forgetting to send the follow-up
- Working at midnight when the lead came in at 11pm
- Scaling without proportional cost increase
Hidden costs nobody warns you about
- Onboarding time: 20-40 hours of your time over the first month writing SOPs and answering questions
- Management overhead: 3-8 hours per week ongoing - reviewing work, redirecting, handling exceptions
- Turnover: average VA tenure is 12-24 months. Every replacement is another full onboarding cycle.
- Scope creep: tasks the VA was hired for grow to fill the available hours, regardless of priority
Path 3: Build an AI automation system
The newer option. Most operators don’t seriously consider it because they think it means custom software or an in-house developer. It doesn’t.
In 2026, a working AI automation system for a small business is built from off-the-shelf tools that already cost you $40-$120 per month combined:
- Make.com or Zapier to orchestrate the workflow ($9-$29/mo)
- ChatGPT or Claude for the AI work ($20-$30/mo per seat)
- Your existing CRM, calendar, and form tools - no new accounts needed
- Jotform or a Google Form for intake ($0-$39/mo)
The build is a one-off. After that, the system runs on autopilot and the only ongoing cost is the tool subscriptions.
What a single 7-day build typically replaces
In the trades and allied health businesses we audit, a single seven-day build sprint usually targets one of these:
- Quote follow-up sequence (industry: trades). A 7-day workflow that fires automatic SMS at 48 hours, 5 days, and 10 days after a quote goes out. Reclaims an estimated 5-15 percent of quoted work that previously went cold.
- No-show reduction (industry: allied health). Cliniko, Power Diary, or Halaxy native reminders plus a custom rebook sequence. Drops no-show rates from 20-30 percent to under 10 percent.
- Review request automation (cross-industry). A trigger 24 hours after job completion or visit. Typical lift: 3-5x the number of Google reviews collected per month.
- Invoice reminder chain (cross-industry). Automated nudges at 7 and 14 days overdue. Typically drops average days-to-payment by 5-10 days.
- Lead intake triage. New enquiries automatically scored, categorised, and routed - so the right person responds first.
What it costs vs what it returns
The one-off build cost varies by who’s building it and what’s being built. A typical seven-day external build engagement in this market sits in the $3,000-$8,000 range. Tools then run $40-$120 per month indefinitely.
At the low end:
- Year-1 cash cost: $3,000 (build) + $480 (tools) = $3,480
- Year-2 cash cost: $480 (tools only)
At the high end:
- Year-1 cash cost: $8,000 (build) + $1,440 (tools) = $9,440
- Year-2 cash cost: $1,440
Compare against a single Australian VA at 15 hours per week running roughly $30,000 per year, every year, forever.
What a system is bad at
- Anything that needs human judgement on every instance
- Anything that requires real-time conversation
- Anything that changes the rules every time
- Anything illegal, ethically grey, or one-off
The honest answer is the same as the homepage answer: most operators end up with both. The system handles the repeatable 80 percent. A VA - usually 5-10 hours per week, not 15 - handles the messy 20 percent.
Side-by-side: 12-month total cost of ownership
For a generic Aussie trades or allied health business with 15 hours per week of admin to offload:
| Scenario | Year-1 cash (AUD) | Year-2 cash (AUD) | Hours/week you spend managing it |
|---|---|---|---|
| Do nothing | $0 + lost revenue | $0 + lost revenue | 0 (you’re still doing the work) |
| Offshore VA (15 hr/wk) | $5,000-$10,000 | $5,000-$10,000 | 3-8 |
| AU VA (15 hr/wk) | $26,000-$37,000 | $26,000-$37,000 | 2-5 |
| AI system (one build) | $3,500-$9,500 | $500-$1,500 | 0-1 |
| System + 5 hr/wk AU VA | $12,500-$22,000 | $9,500-$13,500 | 1-3 |
The bottom row is what most of our audited operators land on after two years. It’s the cheapest version of the answer once the system is paying for itself.
When each path actually wins
Do nothing wins when
- You’re closing, selling, or retiring inside 12 months
- You genuinely have more demand than capacity and don’t care about lost leads
A VA-only path wins when
- Almost all your admin needs human judgement (rare, but real - some allied health practices fit this)
- You don’t run any sequences that repeat the same way (rarer)
- You hate technology more than you hate burning $30k/year (this is most often the real reason)
An AI system wins when
- Most of your admin repeats (quotes, follow-ups, invoices, reminders, review requests)
- You’d rather pay once and own the system than pay forever
- You want capacity that doesn’t increase with cost
System + VA wins when
- You’re a normal operating Australian trades or allied health business
- You’re honest about which 20 percent actually needs a human
That last one is most operators. So the question isn’t “system or VA” - it’s “which one do I build first, and how do I do it without spending $30k on a VA who’s running tasks a $50/month system could handle?”
The honest order: build the system first. Then hire the VA for what the system can’t touch. You’ll need fewer VA hours, and the hours you do buy will go to work that actually needs a person.
Frequently asked
How much does a single AI build sprint cost in Australia?
Pricing varies by builder and scope. A seven-day external engagement targeting a single bottleneck typically sits in the $3,000-$8,000 range in this market. Ongoing tool costs are $40-$120 per month. There are no subscription fees on the build itself - you own the system and the accounts it runs on.
How long until a system pays for itself?
For a build targeting a real revenue leak - quote follow-ups, no-show reduction, invoice chasing - the payback period is typically 30-90 days in trades businesses and 60-120 days in allied health practices. The variability comes from how much revenue was leaking before the build, not how good the build is.
Is it cheaper to learn this and build it myself?
If you bill out at $90+ per hour and you’d need 30-60 hours to learn Make or Zapier well enough to ship something reliable, the maths usually favours buying the build. If you bill out at $30 per hour or you genuinely enjoy the work, building it yourself is reasonable - and the tool subscriptions are the same either way.
What if the system breaks?
Every system we build includes error alerts and a one-toggle kill switch. If the system stops working, you know within minutes, not days. Recovery is usually 15-30 minutes because the build runs on tools you already own - there’s no proprietary platform to debug.
Why not hire a developer to build it?
A small business AI system isn’t a software project - it’s an operations project running on no-code tools. A developer will overbuild it. A specialist operator-builder will ship it in a week using tools your team can maintain.
If you’re sitting on this decision and want a straight answer about which path makes sense for your specific business, the free 30-minute Bottleneck Audit does exactly that. No sales pitch. Plain English. We’ll tell you honestly whether you need a system, a VA, both, or neither.
