There's a version of the AI conversation in advice that goes: this is the one that finally replaces us.
Worth remembering that the profession has run this experiment before. Repeatedly. Since before most currently-licensed advisers were in the industry. And it has never once gone the way the pessimists said it would.
The handoffs
Start in the late 1980s. Advisers ran their own investment administration — application forms, unit registries, distribution reconciliations, CGT schedules, consolidated reporting stitched together by hand. Then master trusts arrived, and then wrap platforms. Navigator. Asgard. Inside a decade, the administration that had defined the back office of a planning practice belonged to someone else.
Then fund selection went. Advisers had built reputations on picking managers. Multi-manager products — MasterKey, Advance, ipac — took the research, the due diligence and the rebalancing, and sold it back as a product. Advisers stopped picking funds and started picking who picks funds.
Then plan preparation went to paraplanners. Then paraplanning went offshore. Then practice administration followed it.
Then portfolio management went. Asset consultants and managed account providers took discretionary control of client portfolios at scale — now a $292.9 billion market, up 25.8% in a single year, with close to three in five advisers using it.
Then underwriting went to insurers' tele-underwriting teams. Then claims went to specialist advocates. Along the way: mortgages to brokers, SMSF administration to specialists, estate documents to lawyers, tax to accountants.
Every one of these was, at the time, called the beginning of the end.
The scoreboard
- Adviser numbers fell from more than 28,000 in 2019 to 15,151 in January 2026
- Median ongoing advice fees rose 93%, from $2,510 to $4,837
- Average practice net margin hit 23.3% — the highest in the four-year series
- The "no profit" cohort nearly halved, from 17.7% in 2023 to 9.4% in 2026
- The average adviser runs $102 million for around 102 ongoing clients
Fewer advisers. Higher fees. Higher margins. More work leaving the desk each year.
Let's be clear about the gold standard
There is nothing better than a skilled adviser sitting with a client for two hours.
Not the questionnaire. Not the software. The adviser — noticing the pause before the answer, asking the question behind the question, catching the thing the client didn't think was relevant and knowing immediately that it's the most relevant thing in the room. Recognising that "we want to retire at 60" actually means "I'm frightened of my job" or "my father died at 67."
No one is going to automate that, and no one should try.
But be equally clear about what it costs. That conversation sits inside an engagement priced at four to six thousand dollars, delivered by a profession of 15,151 people serving around 102 ongoing clients each.
Roughly one in ten Australians receives financial advice. Two-thirds of the unadvised say they'd pay around $500 a year for it. Cost to serve one client is $4,000–$4,500, and mandatory costs alone — licensee fees, PI, ASIC levy — run $36,896 to $83,877 per adviser per year before anyone does a minute of work.
The gold standard is real. It's also, for most of the country, unreachable. That was never an adviser failing. Adviser skill simply doesn't scale, and it has never been affordable to most people.
What everyone else gets instead
Here's the part worth sitting with.
For the client who can't reach the two-hour conversation, the alternative isn't a cheaper adviser. It's a fact-find form.
A PDF. Or a web form with forty fields. Sent with a friendly note asking them to complete it before the meeting.
Ask any practice what happens next. A good number never come back. Others return it half-finished — super fund "not sure," insurance "I think through work," goals reduced to "retire comfortably." No probing, no follow-up question, no explanation of why any of it is being asked. Nothing that shows the client what's actually at stake in their own situation.
A form cannot do the thing a first meeting does. It can't build any sense that this process is worth continuing. Nobody in the history of financial services has read a fact-find form and thought now I understand why I need this.
So the honest comparison isn't AI against an adviser.
It's AI against a form nobody finishes.
Which is the gap worth closing
That's the design brief. Not to replicate what a great adviser does in a room — to give the people who will never be in that room something substantially better than a PDF.
A structured conversation that asks the next question because of the last answer. That explains why it's asking. That notices the gap between what someone says they want and what their numbers can support, and puts that in front of them in plain language while they're still engaged.
There's reason to think this works. In controlled research on screening interviews, participants who believed they were talking to a computer showed lower fear of self-disclosure, less impression management, and were rated by observers as more willing to disclose (Lucas, Gratch, King & Morency, 2014). Not because software is better than a person — because a form offers no reason to be candid, and a conversation, even a software-led one, does.
For the adviser, this cuts two ways, and both are good.
Clients who currently vanish somewhere between enquiry and meeting one arrive instead with a complete picture, a quantified goal, and some sense of why they're there. That's not work taken off the desk. That's clients who never reached the desk at all.
And for full-fee clients, the adviser gets the first hour back. Right now a chunk of the most relationship-critical hour in the whole engagement is spent transcribing super fund names into mandatory fields. Give that time back and it goes where it was always meant to go.
And the value was never the fact-find anyway
The Statement of Advice is the visible artefact, so it carries the fee. The ongoing fee is the one advisers justify every year — because it's the hardest to point at and say here, this is what you paid for.
But the recommendation was always the commodity.
Vanguard's Adviser's Alpha research puts total adviser value at around 3% a year and identifies behavioural coaching as the single largest component, at roughly 150 basis points — bigger than asset allocation, bigger than product selection, bigger than rebalancing.
And here's the twist: when Morningstar surveyed Australian clients and advisers, clients ranked technical attributes above behavioural ones. Clients don't consciously value the thing that delivers them the most value.
Because goals set on day one have a shelf life. Divorce. An inheritance. Redundancy at 58. A diagnosis. A child who doesn't launch. A business that fails.
The plan is a snapshot of a moment that stops being true almost immediately. The value isn't the plan. It's March 2020. It's the 7pm call after the retrenchment. It's the client who wants to go to cash and needs someone with the standing to say no — and to be accountable if they're wrong.
A model can generate those words. It cannot occupy that position.
"But this time is different"
Fair. Every previous handoff moved labour. This one moves reasoning, and advisers are right to look harder at it.
Except ASIC has already reviewed offshore outsourcing by advice licensees, where the main functions going offshore are paraplanning and advice support. Its concerns are control, data protection, confidentiality and breach detection. They are not about whether someone other than the adviser may draft the advice. That was settled decades ago. The licensee remains accountable regardless of who holds the keyboard.
So the bar isn't "can it write an SoA." It's whether every output is reviewed by a human before it reaches a client, whether every recommendation traces back to the input that drove it, and whether the file survives an AFCA determination three years from now. Anything else is a liability, not a tool.
The pattern holds
3.59 million Australians over 65 have no adviser. Projections say the country needs 50,000+ advisers by 2055. There are 15,151.
You can't recruit out of that. You can't offshore out of it.
Thirty-five years, one repeated move: hand off the production, keep the judgment, serve more people. The fact-find was always production. The relationship that catches a client when their life comes apart never was.
The best experience is still an adviser in a room. For everyone who can't get there, the choice today is a form nobody finishes — and that's the gap worth building for.