In mortgage lending, time kills deals. New broker feedback indicates assessment queues stretching across the market even as applications climb — a textbook capacity crunch that risks revenue leakage and reputational damage. Historically, aggregator reporting has shown wide gaps in turnaround times between majors and non‑majors; when volumes surge, those differences widen. The competitive stakes are clear: for brokers, reliability often trumps a few basis points of rate. For banks, the cost of delay compounds quickly.

Operations lens: Why queues blow out fast

Two fundamentals explain today’s pain. First, Little’s Law: lead time equals work-in-progress divided by throughput. If application inflow rises and underwriting capacity stays flat, work-in-progress swells and lead times stretch. Second, Kingman’s formula shows delays expand exponentially with variability. In lending, variability comes from mixed customer profiles (PAYG versus self-employed), document completeness, valuation complexity and policy exceptions. Volatility in any of these — combined with fixed capacity — turns a busy day into a weeks-long backlog.

Compounding the issue, process handoffs (broker → credit assessor → valuations → loan documentation → settlement) act like serial queues. A thin spot in any node creates system-wide latency. When brokers observe turnaround ‘blowouts’, the root cause is almost always an unmanaged constraint — usually manual underwriting or valuation scheduling — amplified by uneven work arrival and inconsistent triage.

Business impact: The silent P&L hit from slow decisions

The financial drag from slow assessment is underappreciated. Consider an illustrative scenario: a mid-sized lender receives 2,000 applications per month. If queue-induced churn causes even 10 per cent of otherwise approvable customers to switch mid-process, and if each settled loan delivers a modest gross margin (for example, basis points on average loan size), the lost contribution can run into millions annually. On top of direct revenue, there is acquisition waste (marketing and broker commission without settlement), higher rework costs as documents expire, and reputational damage that reroutes brokers to competitors for months.

Broker behaviour reinforces the penalty. Aggregator data has previously shown brokers rebalance flows toward lenders with consistent service-levels when volumes tighten. Once a lender drops to the bottom of a broker’s recommendation set, recovery takes time — even after capacity returns — because trust lags. In short, turnaround is not a back-office metric; it is market share in disguise.

Competitive dynamics: Service-level as strategy

Lenders have three strategic choices for speed: dominate, segment or price it. Dominators make turnaround a franchise attribute — think tight underwriting rules, predictable SLAs, and disciplined broker communication. Segmenters build dual lanes: a fast track for low-risk, high-data-density applications (e.g., PAYG with clean credit) and an expert lane for complex deals. Pricers explicitly monetise speed, offering priority channels for a fee or via tighter pricing bands, while maintaining baseline service for everyone else.

Non-banks often move quicker because they can refactor policy and workflow without legacy constraints, but funding costs can narrow their pricing room. Majors, meanwhile, can invest in tooling and workforce depth at meaningful scale. For brokers, the calculus is simple: reliable 48–72 hour initial credit decisions win flow. For lenders, published and kept SLAs that are visible in broker portals become a competitive weapon.

Technology and data: The new speed stack

Operational speed now relies on a distinct technology stack that reduces variability and manual touch. Core components include:

• Data ingestion and prefill via Consumer Data Right (CDR): With customer consent, lenders can pull verified transaction and account data from source, cutting document chase and errors. That reduces work-in-progress and increases first-time-right rates.
• Document intelligence: OCR plus classification and entity extraction to auto-recognise payslips, BAS statements and ID documents. Straight-through validation for standard artefacts reduces underwriter load.
• AI triage and risk scoring: Machine learning models route files by complexity, flag policy exceptions early, and suggest required documents. In Australia, deployments should align to the government’s AI Ethics Principles that call for systems that are “safe, secure and reliable,” with clear human oversight.
• Valuation orchestration: API-based ordering, automatic selection of desktop versus full valuation based on risk rules, and dynamic slotting to available valuers to remove a frequent bottleneck.
• Workflow visibility: Real-time SLA dashboards to brokers, with expected decision times, missing information alerts, and escalation paths.

The Australian Taxation Office’s governance discussion on general-purpose AI underscores the need for explainability and auditability in public institutions; lenders should adopt the same posture. Black-box models that cannot explain declines will not survive scrutiny from credit risk, regulators or customers.

Implementation reality: A 90–180 day playbook

Speed improvements do not require a core replacement. A practical rollout looks like this:

• Map the constraint: Measure end-to-end lead time, queue lengths by stage, rework rates and “first-time-right” documentation. Publish one truth to executives and brokers.
• Segment and triage: Define low-complexity criteria for straight-through pre-approval. Build a “green lane” with tight policy and a “red lane” for human-led expert assessment.
• Backlog burn-down: Ring-fence a surge squad (contract assessors, overtime, weekend shifts) to reset the system. Without a burn-down, new tools drown in old work.
• Automate the first hour: Introduce document capture, CDR consent and automated completeness checks at lodgement to reduce variability and underwriter touches.
• Valuation and settlement orchestration: Integrate ordering, status updates and slot optimisation; pre-book settlements once conditions precedent drop below a threshold.
• Broker communication: Provide time-stamped SLAs at file level; proactively flag missing items; allow one-click escalation with reason codes.
• Control and compliance: Embed AI governance guardrails — model risk reviews, bias checks and human-in-the-loop controls — consistent with national principles.

Critically, measure outcomes that matter: time to initial credit decision, percentage of files decided within SLA, straight-through pre-approvals, and broker NPS. Celebrate speed wins publicly; brokers reward predictable performance.

Market context and outlook

Australian borrower activity is cyclical and sensitive to rates, sentiment and migration. When demand turns, capacity often lags by quarters because lenders staff cautiously. Expect continued volatility in volumes over the next year; the winners will design for elasticity: cross-trained assessment pods, overflow partnerships, and scalable cloud-based tooling. Aggregators will likely amplify transparency on service levels, further rewarding operationally disciplined lenders.

One adjacent lesson: in other digital markets, dominant players win by compressing decision time at meaningful scale. The ACCC notes Google holds roughly 94 per cent search share in Australia (August 2024) — a reminder that responsiveness and reliability compound. Mortgages are slower and more regulated, but the principle holds: customers and intermediaries cluster around systems that feel instant and certain.

What leadership should do now

Boards and executives should treat turnaround as a strategic KPI, not an operational afterthought. Approve a short, focused investment in the speed stack, authorise temporary capacity to clear backlogs, and publish broker-facing SLAs you can keep. Align AI-enabled triage with national ethics guidance and your credit risk appetite. Finally, incentivise speed: tie leader bonuses to time-to-decision improvements and settlement conversion, not just volume. In a rising-application market, speed-to-yes is the cheapest form of growth capital you can buy.