Dan Siegler’s call for Australia’s brokers to dismantle legacy practices lands at a pivotal moment. The broker channel already intermediates the majority of new mortgages, yet revenue pools are under pressure and the customer experience is uneven. With artificial intelligence accelerating in the United States and UK markets, and Australia’s own policy moves reshaping demand, the sector faces a choice: modernise the operating model or drift into price-based competition.

The signal behind the noise: a strategy read on disruption

Look past the hype and apply two lenses. First, the industry S-curve: digital workflows in origination have crossed from early adopters to mainstream lenders; the next curve is AI-augmented advice and straight-through processing. Second, Five Forces: buyer power is rising as price/feature transparency improves; substitutes are emerging in direct-to-consumer fintech and embedded finance; the threat of new entrants is increasing as proptechs and digital advisors leverage low-cost stacks. In this context, Siegler’s message is not about gadgets — it’s about re-architecting the value chain.

Market context: growth, policy and regulation reshape incentives

By 2025, estimates of Australia’s mortgage broking market revenue range between roughly AU$4.2 billion and AU$6.2 billion, reflecting methodological differences but pointing to a meaningful, contested pool. Brokers originate roughly two-thirds to three-quarters of new home loans in recent periods, underscoring systemic importance. Two macro shifts now bite:

• Policy demand shocks: The accelerated expansion of the Home Guarantee Scheme (HGS) from 1 October 2025 aims to bring more first-home buyers into the market. That can lift volumes but also risks stoking entry-level price inflation, dampening affordability gains and forcing brokers to manage client expectations and lender capacity constraints simultaneously.
• Duty of care and compliance cost: The Best Interests Duty (BID) has strengthened consumer trust but requires richer product comparisons, clearer rationale capture and auditable advice trails. Firms that automate compliance will carry structurally lower unit costs — and higher resilience — than those relying on manual processes.

Technical deep dive: the modern broker tech stack

Disruption here is a capability rebuild, not a point-solution spree. A pragmatic stack looks like this:

• Data ingestion and normalisation: Bank statements, payslips and IDs processed via OCR and computer vision, with entity resolution to create a clean customer file. This removes rekeying and reduces error rates.
• Document intelligence and eligibility engines: NLP models extract income, liabilities and anomalies; policy rules engines map borrower profiles to lender criteria in real time, flagging showstoppers early.
• Property intelligence: Automated valuation models (AVMs), comparable sales engines and risk overlays (LVR, postcode concentration, flood/fire risk) provide valuation ranges and risk flags during pre-qualification.
• AI co-pilots for advisors: Large language models, grounded on internal knowledge bases and lender policy documents, summarise options, draft BID-compliant recommendations, and generate client-ready explanations in plain English — with human-in-the-loop approval.
• Workflow and straight-through processing: API-first application orchestration with lenders, eIDV, eSign and eSettlement to compress time-to-approval and reduce fallout.
• Analytics and model risk: Dashboards on time-to-yes, abandonment, approval conversion, clawbacks and arrears cohorts; MRM pipelines to monitor drift and bias, aligned with privacy and fairness requirements.

The technology is available off-the-shelf from vendors across fintech, regtech and proptech. The differentiator will be data quality, process design and change management.

Economics that matter: from cost-to-serve to conversion

Three levers determine whether technology investments pay back:

• Cycle time: Every day shaved off time-to-yes reduces leakage to competitors and improves NPS. In markets where AI triage and document intelligence have been deployed, firms report materially faster pre-assessment and fewer back-and-forths with lenders. Faster processing directly lifts conversion and referral rates.
• Right-first-time submissions: Eligibility engines and policy co-pilots cut rework. Fewer referral declines reduce adviser time lost, clawback risks and client frustration.
• Compliance at meaningful scale: Automated file notes, advice rationale and product comparison evidence lower the marginal cost of BID while raising audit readiness. That is a structural margin advantage, not a cosmetic win.

Competitive edge: where boutiques and incumbents can win

Nimble, specialised firms are well positioned to outmanoeuvre larger brands in niches — self-employed borrowers, expats, complex income, or regional markets — by tuning data models and workflows to local realities. Meanwhile, incumbents can leverage brand trust and distribution breadth if they industrialise advice, deploy personalisation at meaningful scale and build data partnerships.

Three strategies stand out:

• Specialisation and playbooks: Build domain-specific policy graphs (e.g., for construction loans or asset finance) and publish transparent service-level commitments that beat the market.
• Partnerships and embedded origination: Co-create pre-qualification experiences with real estate portals, developers and employers. Embedded finance funnels convert higher and lower acquisition costs than generic lead buying.
• Data moat creation: Invest in consents-driven data capture (open banking, payroll APIs) and feedback loops (post-settlement performance) to improve matching and risk screening over time.

Implementation reality: governance, ethics and skills

Adopting AI is as much about controls as capability:

• Privacy and security: Align with Australian Privacy Principles; prefer on-shore processing and encryption at rest/in transit; maintain vendor SOC 2/ISO 27001 evidence; use data minimisation and de-identification for model training.
• BID-aligned AI: Ground LLMs on approved content; disable free-form hallucination for advice; enforce human approval gates; capture rationale and version every output for audit.
• Model risk management: Set thresholds for accuracy and bias; monitor drift; implement incident response for model failures; keep fallback manual pathways.
• People and change: Upskill advisers as “augmented analysts”; redesign incentives to reward quality and compliance; appoint a product owner for the broker platform to avoid tool sprawl.

Policy ripple effects: HGS and the affordability paradox

The HGS expansion will likely lift enquiry volumes among first-home buyers but could add heat to price-sensitive segments. For brokers, this means stress-testing serviceability at realistic buffers, setting client expectations on competition for stock, and coordinating with lenders on capacity constraints. Transparent, data-backed guidance will be a differentiator when emotions run high.

What good looks like in 12 months

Leading firms will show measurable gains: 20–30% faster pre-assessment, material lift in right-first-time submissions, better audit outcomes, and higher cross-sell into protection and asset finance — all while keeping complaint rates low. The laggards will still be pushing PDFs around and wondering why referral partners are drifting.

The next 24–36 months: consolidation and new business models

Expect three shifts: consolidation around data-rich platforms; growth of subscription-style service tiers for ongoing credit health monitoring; and tighter lender-broker API standards enabling near real-time decisioning. Those who treat AI as the new operating system for distribution — not a bolt-on — will define the next chapter.

Action plan for principals and CEOs

• Map the client journey; quantify drop-offs and rework; prioritise two bottlenecks for elimination.
• Stand up a secure data layer; pilot document AI and an eligibility engine in parallel; measure right-first-time and cycle-time deltas.
• Codify BID into workflows; automate rationale capture; rehearse for audit.
• Strike one embedded origination partnership; test CAC and conversion economics against traditional channels.
• Institute model risk governance; name an accountable AI owner; train advisers as co-pilot users.

The strategic takeaway: technology is merely the enabler. The real disruption is an operating model that is faster, more transparent and compliance-native. Build that — or prepare to compete on price alone.