Back to Blog Industry 7 min read

The ROI of Embedded Finance: How SaaS Platforms Are Calculating the Business Case

Adding payments or lending to your SaaS platform can increase ARPU by 2–4x. Here's how to build the business case for your leadership team.

LP
Laura Petrov
Embedded finance ARPU increase chart for SaaS platforms

The pitch for embedded finance is compelling in the abstract: add payments or lending to your SaaS platform, capture a share of the financial flows your customers are already running, and meaningfully increase ARPU without acquiring new customers. The numbers that get quoted — "2–4x ARPU increase," "customers using embedded finance churn 30–50% less" — are directionally real. But they're also aggregate figures that smooth over significant variance across platforms and implementation approaches.

If you're building the business case for your leadership team or investors, aggregate industry estimates aren't sufficient. You need a model that works through your specific customer economics, implementation costs, and realistic adoption curves. This article walks through how to structure that analysis.

Revenue model: where the money comes from

Embedded finance revenue for a SaaS platform comes from several sources, and they don't all arrive at the same time or at the same margin.

Payment interchange

When your platform processes card payments on behalf of your customers or their end users, you can participate in interchange revenue — the fee the card network charges on each transaction. For Visa/Mastercard credit transactions, interchange rates typically run 1.5–2.5% of transaction value. Debit interchange under the Durbin Amendment for large issuers is capped at $0.21 + 0.05% per transaction, but for platforms using a bank partner not subject to Durbin, rates can be higher.

Interchange is a volume game. A field service SaaS with 500 customers each processing an average of $50,000/month in service payments would represent $25M/month in GMV. At 1.5% interchange and a 30% platform share of that interchange, the monthly revenue contribution is $112,500 — a meaningful number. The caveat: interchange sharing arrangements require a bank sponsorship relationship and the economics depend heavily on how much of the interchange the bank retains.

Net interest margin on embedded lending

If your platform is acting as program manager for a lending product — where a bank partner funds the loans — the platform typically captures a portion of the interest spread. Small business loan rates through embedded programs generally run 8–18% APR. After the cost of funds to the bank (currently 5–6% for funded loans), the net interest spread available is roughly 3–12 percentage points, of which the platform might capture 2–5 percentage points depending on the program agreement.

On a $10M outstanding loan portfolio with a 3% platform spread, that's $300,000/year in interest revenue. Building to that portfolio size from zero takes time — typically 18–36 months from launch depending on adoption rates and average loan size.

Late fees, float income, and ancillary revenue

Lending programs generate fee income beyond interest: origination fees (0.5–2% of loan amount, collected at disbursement), late payment fees, and in some cases prepayment fees. These are typically modest individually but accrue at volume. Float income — interest earned on funds held in transit during payment processing — has become more significant in higher interest rate environments, though it depends on program structure and banking partner arrangements.

Cost model: what actually gets capitalized

The cost model is where most embedded finance business cases become unrealistic. Teams routinely undercount the non-engineering costs.

Engineering time

If you're using an API-first embedded finance provider, the integration engineering is manageable: typically 2–6 weeks of focused engineering work for a payments integration, 6–12 weeks for a lending integration with custom underwriting logic. If you're building direct bank connections or custom compliance infrastructure, add a full team for 6–12 months.

Compliance infrastructure

This is where the cost model most often breaks. A minimum viable compliance setup for a lending program includes: KYC/KYB identity verification (typically $1–3 per check at volume), BSA/AML transaction monitoring software ($30,000–$100,000/year for a SaaS-appropriate solution), and at least part-time compliance oversight. If you're hiring in-house, a compliance manager with fintech experience runs $120,000–$180,000 in base salary in San Francisco. Compliance counsel on retainer adds $50,000–$100,000/year for ongoing guidance.

For a platform using a compliance wrapper that bundles these functions, the cost structure shifts to a per-transaction or flat fee model — often significantly cheaper at early volumes than the DIY stack. The crossover point (where DIY becomes cheaper than a wrapper) typically occurs at transaction volumes well above where most platforms launch.

Customer support and operational overhead

Financial products generate a different category of customer support than SaaS products. Payment disputes, loan servicing questions, failed transaction inquiries — these require support staff with financial product knowledge. Budget for incremental support headcount or specialized outsourcing as volume grows.

NPV framework: a worked example

Consider a vertical SaaS platform serving 500 SMB customers, each paying $200/month for the core software product ($100,000/month ARR, $1.2M ARR). The platform adds embedded payments and a working capital lending product.

Adoption assumptions (conservative): 35% of customers adopt payments (175 customers), 15% adopt lending (75 customers). These are deliberate underestimates — actual adoption rates in well-integrated embedded finance programs typically run higher, but starting conservative protects the business case from being dismissed as optimistic.

Revenue Stream Monthly (Steady State) Annual
Payments interchange (175 customers × $40K GMV × 0.5% platform share) $35,000 $420,000
Lending interest spread (75 customers, $15K avg outstanding, 3% spread) $2,813 $33,750
Loan origination fees (avg 1% × $15K × 75 customers, annualized) $11,250
Total embedded finance revenue ~$38,000 ~$465,000

Against an existing $1.2M ARR base, $465K in incremental annual embedded finance revenue represents a 39% revenue uplift — consistent with the "add $80/month embedded finance ARPU" thesis at 175 customers using payments. The cost to achieve it (compliance wrapper fees, incremental support, integration engineering) runs $80,000–$150,000 in year one, declining in subsequent years as the setup costs amortize.

Churn reduction: the harder-to-model but real effect

The churn reduction case is genuinely difficult to model in advance, because it depends on how deeply integrated the financial product becomes in customer workflows. The pattern that drives retention is workflow lock-in: a customer who processes payroll and runs working capital loans through your platform has two workflows inside your software that would be painful to migrate. The more financial workflows are entangled with operational workflows, the higher the switching cost.

Industry-range estimates for churn reduction among customers using embedded financial products run 20–50% lower than the platform's base churn rate. At a 10% annual base churn rate on 500 customers (50 churns/year, costing ~$600K ARR at $12K average contract), even a 25% churn reduction (12–13 fewer churns/year) preserves $150,000 in ARR annually. That's a meaningful line in the NPV model even before you count direct financial product revenue.

We're not saying churn reduction alone justifies an embedded finance buildout. We're saying it's a legitimate line item in the NPV model that often gets omitted because it's harder to attribute causally. A reasonable approach: use a conservative 20% churn reduction estimate for customers actively using financial products, and build sensitivity analysis around the range.

When the business case is weak

Not every vertical SaaS platform should build embedded finance. The business case is weakest when:

  • Your customers' transaction volumes are low (sub-$5,000/month average). The interchange revenue at that volume is not worth the operational overhead.
  • Your customers don't have a payment or financing need that connects to your core workflow. Forcing a financial product where the workflow connection is weak produces low adoption regardless of how well the product is built.
  • Your platform is pre-product-market fit on the core SaaS offering. Embedded finance adds compliance and operational complexity that will compete for engineering attention. Get the core right first.

The business case math here is real for the right profiles. The honest version of the analysis includes realistic adoption assumptions, full cost accounting (especially compliance infrastructure), and a multi-year time horizon — steady-state embedded finance economics typically take 18–24 months to materialize after launch.