Context: The demand spike meets a distribution shift

Small and medium enterprises power Australia’s economy, comprising about 98% of all businesses. Yet working capital and growth finance remain stubborn chokepoints. The International Finance Corporation has long estimated an annual global SME credit gap exceeding US$5 trillion, and local data echo the theme: Australia’s SME lending stock was estimated at approximately $631 billion in February 2024, expanding at a 9.4% compound annual rate since 2019. The fuel? A post-pandemic rebuild, inventory swings, supply-chain reconfiguration, and digitisation projects that can’t be funded off a credit card.

Distribution is changing just as fast as demand. Fintechs have redesigned origination and underwriting for speed, banks are modernising legacy processes, and embedded finance has turned software platforms into lenders. Industry analysts, including Riverview Finserve, call out three reinforcing trends for 2025: the rise of alternative lenders, embedded finance inside business platforms, and hyper-personalised offers powered by AI and richer data.

Decision: Two paths that define the playbook

This case study focuses on two real-world strategies that leaders are using to win SME lending profitably:

Case A — Specialist bank scaling relationship-led SME lending (Australia): Judo Bank built an SME-only franchise from scratch, pairing experienced bankers with modern tooling. The decision was to reject universal banking economics (high-volume, commoditised credit) and instead compete on speed, access, and human judgement where the data are messier than consumer credit — and the margins better.

Case B — Embedded lenders inside business platforms (global): Shopify Capital and Square Loans (Block) integrated lending into merchants’ daily workflows, using checkout, sales, and settlement data to pre-score offers. The strategic decision was to prioritise distribution and data moats over storefront marketing. Offers appear contextually; repayment aligns with cash flow via revenue share.

Implementation: The technology and operating model that make it work

Data advantage: SME underwriting hinges on cash-flow visibility rather than collateral alone. Leaders ingest bank feeds (via Open Banking/Consumer Data Right in Australia), accounting ledgers, e-commerce carts, and payment processors. Signal-rich features include sales volatility, invoice ageing, seasonality, and supplier concentration. The goal: continuous, permissioned telemetry to turn lumpy SME risk into a managed, dynamic portfolio.

Decisioning stack: Successful lenders combine:

• Rules for eligibility and policy guardrails (sector exposure, loan-to-revenue thresholds, director checks).
• Machine learning models for probability of default and loss given default, trained on transaction-level histories to reduce thin-file penalties.
• Graph/network analytics to detect related-entity risk and fraud rings.
• Human-in-the-loop workflows for exceptions, complex financials, and covenant setting.

Orchestration: Digital onboarding (eKYC/KYB), document ingestion with OCR, and straight-through processing for lower-risk tranches. Cloud-native core lending systems and API gateways connect brokers, bank RMs, and platform partners. For brokers, lender marketplaces and decisioning portals standardise product comparison and term sheets — compressing time-to-yes from weeks to hours in many cases.

Compliance and capital: Banks align with APRA prudential standards and Basel risk weights; all lenders must meet AML/CTF obligations and Privacy/CDR rules. Non-bank fintechs compete with speed and UX, then secure scalable funding through warehouse lines, securitisation, or bank partnerships to lower cost of funds — a key determinant of sustainable pricing.

Results: What the leaders achieved (with numbers)

Market outcomes: Australia’s SME lending pool is estimated at ~$631 billion (Feb 2024), growing at about 9.4% CAGR since 2019 — a strong tailwind for brokers diversifying beyond residential mortgages. Globally, the SME financing shortfall remains above US$5 trillion annually, underscoring runway for alternative models.

Case A — Judo Bank (Australia): By focusing exclusively on SMEs, pairing relationship bankers with modern credit tools, and competing on time-to-credit, Judo scaled its loan book to around the A$10 billion mark by 2024. The bank’s proposition — banker access, cash-flow lending, and faster decisions — translated into durable customer acquisition without mass-market marketing spend.

Case B — Embedded lenders (global): Shopify Capital and Square Loans have each originated billions of dollars to SMEs, with approvals often delivered within minutes and funds landing as fast as the next business day. Repayments tied to daily sales reduce friction and improve affordability, helping sustain repeat borrowing and lower acquisition costs through native distribution.

Operational KPIs seen in high performers:

• Time-to-yes routinely compressed from 10–20 business days to under 24–72 hours for standard cases.
• Acquisition costs fall materially when distribution is embedded (e.g., within commerce/payments platforms) versus direct marketing.
• Portfolio monitoring moves from monthly to near real-time, enabling faster risk interventions and loss containment.

Market context and competitive dynamics

Porter’s view: Threat of new entrants stays elevated as cloud cores and Banking-as-a-Service lower setup costs. Buyer power (SMEs) is rising because switching costs fall when offers are instant and portable. Substitutes — including BNPL-for-business and merchant cash advances — keep pricing disciplined. Incumbent banks’ advantage remains cost of funds and trust; fintechs win on speed, data, and UX. Expect more partnerships, white-labelling, and co-origination.

Trendlines to monitor:

• Embedded finance: Software platforms, marketplaces, and PSPs continue integrating credit at the point of need.
• AI underwriting: From scorecards to adaptive models and explainability tooling that satisfies regulators and auditors.
• Open Banking/CDR maturity: Better data granularity and consent flows unlock richer cash-flow analytics and faster drawdowns.

Implementation reality: How brokers and lenders execute

For mortgage brokers diversifying into SME:

• Build a segmented product shelf (working capital, asset finance, invoice finance, merchant cash advances) and map to industries you already serve.
• Invest in business financial literacy — train advisers to read cash-flow statements and spot lending triggers (e.g., inventory bulges, receivables ageing).
• Use multi-lender SME platforms with real-time data pulls (bank feeds, accounting packages) to pre-qualify clients in the first meeting.
• Create post-settlement programmes: quarterly check-ins, covenant monitoring alerts, and proactive top-up offers.

For lenders:

• Prioritise data partnerships (accounting, commerce, payments) to reduce thin-file risk and enable pre-approved offers.
• Engineer an explainable AI pipeline: feature stores, bias monitoring, challenger models, and human sign-off tiers.
• Lower cost of funds early via committed warehouses and bank partnerships; unit economics depend on spread minus expected loss minus acquisition cost.
• Institutionalise “second line” risk: model risk management, stress testing by sector, and early-warning systems using anomaly detection.

Lessons: What decision-makers should take away

First, SME lending is no side hustle: the addressable pool is large and growing, and the economics can be superior to commoditised home loans if you solve for speed, insight, and service. Second, distribution is destiny: embedded and broker-led channels outperform cold direct acquisition. Third, data beats collateral when it comes to underwriting underserved businesses; cash-flow telemetry plus explainable models are now table stakes. Fourth, partnership beats purity — banks, fintechs, and platforms will increasingly co-originate to balance speed, risk, and cost of funds. Finally, governance matters: get AML/CTF, Privacy/CDR, and model risk right early to avoid growth-without-control — the fastest route to value destruction.

In short: the lenders and brokers who treat SME credit as a core business, not a bolt-on, will compound advantage as the cycle turns. Those who hesitate will watch their customer relationships — and margin — migrate to whoever shows up first with capital precisely when it’s needed.