How QuickBooks Capital's $4.3B Lending Surge Reshapes AI Underwriting for Merchant Cash Advance

Platform Lenders Are Raising the Bar for Every MCA Funder

QuickBooks Capital just posted another massive quarter. Intuit's lending arm originated approximately $1.7B in small business loans during Q3 of its fiscal year 2026, pushing the trailing nine-month total to roughly $4.3B. During the earnings call, CEO Sasan Goodarzi doubled down on a message he has been delivering for quarters: AI is not a threat to Intuit. It is an advantage.

That statement should land differently if you run an independent MCA operation. QuickBooks Capital does not compete on brand recognition or marketing spend. It competes on data proximity. Every invoice, every categorized transaction, every bank reconciliation that a merchant performs inside QuickBooks becomes a pre-underwritten signal. By the time a merchant clicks "apply," the platform already knows their average daily balance, revenue trend, and cash flow volatility. The underwriting decision is not instantaneous by accident. It is instantaneous because the data pipeline started months before the application.

For MCA funders who still collect bank statements via email, chase missing pages, and key data into spreadsheets, this is a widening competitive gap. The question is not whether AI underwriting for merchant cash advance will become standard. It already is standard for the largest originators. The question is how quickly smaller funders can build a comparable data pipeline without owning the merchant's accounting software.

Why Data Proximity Is the Real Moat, and How to Build Your Own

The Embedded Data Advantage

Platform lenders like QuickBooks Capital, Square, and Shopify enjoy what the industry calls "embedded data advantage." They see the merchant's full financial picture in real time because the merchant already uses their software for daily operations. Underwriting becomes a byproduct of the platform relationship rather than a separate, friction-heavy process.

Consider what this means in practice. A QuickBooks Capital applicant does not need to download three months of bank statements, convert them to PDF, and upload them through a portal. The platform already has categorized transaction data, receivables aging, and payroll obligations. AI models trained on millions of similar merchant profiles can score the application in seconds. As we explored in our analysis of how QuickBooks Capital's data moat reveals the future of AI underwriting for MCA, this structural advantage compounds over time as the models ingest more data and refine their predictions.

Square's lending arm tells a similar story. Block's Q1 2026 results showed Square Loans driving gross profit growth, with estimated originations near $1.9B for the quarter. These platforms do not just lend faster. They lend with better default predictions because they see the merchant's revenue in real time, not through a static PDF snapshot.

Closing the Gap Without Owning a Platform

Independent MCA funders will never own the merchant's accounting software. That is a reality worth accepting rather than lamenting. But the gap between platform lenders and traditional funders is not really about owning the software. It is about the speed and accuracy of data ingestion. If you can get from "application received" to "underwriting-ready data" in minutes instead of days, you recover most of the speed advantage that platforms enjoy.

This is where the technology stack matters enormously. AI-powered bank statement extraction can parse uploaded PDFs and pull out average daily balances, deposit patterns, NSF counts, and ending balances without any manual data entry. Document classification models can sort a mixed upload of bank statements, tax returns, driver's licenses, and voided checks into the correct categories automatically. Transaction categorization algorithms can flag revenue deposits versus transfers versus loan payments, giving underwriters a clean cash flow picture comparable to what a platform lender sees natively.

The remaining bottleneck, and arguably the biggest one, is the applicant intake process itself. If the merchant has to email documents, wait for a confirmation, get asked for missing pages, and email again, the speed advantage of AI extraction evaporates. Async intake solves this. A single upload link, sent once, lets the merchant drag and drop every document at their convenience. AI extraction begins the moment files land. By the time an underwriter opens the deal, structured data is already waiting for review.

AI as a Competitive Weapon, Not a Feature Checkbox

Goodarzi's framing of AI as an advantage rather than a threat is worth unpacking. He is not talking about chatbots or marketing copy generators. He is talking about AI models that sit inside the lending decisioning engine, continuously learning from repayment outcomes, adjusting risk scores, and identifying patterns that human underwriters miss.

For MCA funders, the actionable version of this looks like several specific capabilities. First, anomaly detection in bank statements: AI models that flag when deposit patterns suddenly change, when round-number deposits suggest cash infusions designed to inflate revenue, or when statement formatting inconsistencies suggest document manipulation. Second, predictive default scoring that goes beyond simple averages to model cash flow volatility, seasonal patterns, and the merchant's position in their industry cycle. Third, stacking detection that identifies existing MCA positions by recognizing daily or weekly debit patterns consistent with existing advances.

These are not hypothetical capabilities. They are what the leading platforms already deploy at scale. The difference is that independent funders need to assemble these capabilities from best-of-breed tools rather than building everything in-house. As we discussed in our piece on how AI-driven financial decision tools are reshaping bank verification software for funders, the build-versus-buy calculus has shifted dramatically as purpose-built AI tools have dropped in cost while gaining accuracy.

What This Means for Your Underwriting Pipeline Today

Let's ground this in a scenario that most MCA operations will recognize. A broker submits a deal at 2 PM. The application PDF is attached, along with four months of bank statements from two different accounts and a copy of the merchant's driver's license. In a manual workflow, someone on your team opens the email, downloads the attachments, names the files, enters basic business information into your CRM, and begins reviewing statements page by page. If a page is missing or illegible, they email the broker, who emails the merchant, who may or may not respond before end of day. Realistic turnaround to underwriting-ready status: 4 to 24 hours.

Now consider the same deal flowing through an AI-powered async pipeline. The broker forwards the email to a dedicated inbox. Document classification identifies each file type. AI extraction pulls business name, EIN, owner details, and bank account information from the application. Bank statement OCR captures monthly deposits, withdrawals, average daily balances, and ending balances. Within minutes, the deal appears on the underwriter's dashboard with structured data already populated and flagged for review. Missing documents trigger an automatic upload link sent to the merchant. No phone calls. No email chains.

The time saved is meaningful, but the real value is competitive. According to the Federal Reserve's Small Business Credit Survey, speed of decision is one of the top reasons small businesses choose alternative lenders over banks. When QuickBooks Capital can approve a line of credit in under a minute, every hour you add to your process is a merchant you risk losing. MCA adoption has climbed to 7% of small businesses on a regular basis, up from 6% the prior year. The market is growing, but the merchants entering it expect platform-speed experiences even from independent funders.

This is precisely the workflow Let's Submit was designed to enable. One link for document collection, AI-powered extraction the moment files arrive, and a dashboard that gives underwriters structured data instead of raw PDFs. The platform does not replace underwriter judgment. It eliminates the hours of intake work that prevent underwriters from exercising that judgment quickly.

The competitive landscape reinforces this urgency. Square's lending growth, Shopify's $1.4B MCA quarter, and now QuickBooks Capital's $4.3B nine-month run rate all point in the same direction: platform lenders are scaling originations at rates that would have seemed impossible five years ago. Independent funders who match their speed on intake and data structuring remain competitive. Those who do not will find themselves competing for the deals that platforms decline, which is a viable but increasingly narrow lane.

Frequently Asked Questions

What is AI underwriting for merchant cash advance?

AI underwriting for merchant cash advance refers to using machine learning models and automated data extraction to evaluate a merchant's cash flow, revenue patterns, and risk profile without relying solely on manual document review. In practice, this includes AI-powered bank statement parsing that extracts deposit totals and average daily balances, anomaly detection that flags inconsistencies in financial documents, and predictive scoring models that estimate repayment probability based on historical patterns. Platform lenders like QuickBooks Capital and Square use embedded AI underwriting natively. Independent MCA funders can achieve similar capabilities by deploying specialized tools for document extraction and bank verification.

How do platform lenders like QuickBooks Capital underwrite so much faster than traditional MCA funders?

Platform lenders underwrite faster because they already possess the merchant's financial data before the application is submitted. A QuickBooks user's invoices, expenses, and bank connections are continuously updated within the platform, so the lending engine can score creditworthiness in real time without requesting external documents. Traditional MCA funders, by contrast, must collect bank statements, verify their authenticity, and manually extract key metrics before underwriting can begin. Closing this gap requires automating the document intake and extraction process so that data reaches the underwriter in structured form within minutes of submission.

Can independent MCA funders compete with platform lenders on speed?

Yes, but it requires deliberate investment in the data pipeline. Independent funders will not replicate the embedded data advantage of a QuickBooks or Square, but they can dramatically reduce intake time by using AI-powered document classification, automated bank statement OCR, and asynchronous applicant upload portals. The key is removing manual steps between "documents received" and "underwriter reviews structured data." Tools like Let's Submit compress this window by providing a single upload link for applicants and running AI extraction automatically, delivering underwriting-ready data to the dashboard without manual data entry. As covered in our discussion of how speed to lead depends on bank verification software for funders, the funders who automate intake consistently win more deals.

Is AI replacing human underwriters in MCA lending?

Not in any meaningful sense for independent MCA operations. AI excels at structured data extraction, pattern recognition, and flagging anomalies, but the final funding decision still requires human judgment, especially for edge cases involving industry-specific risk, stacking concerns, or unusual cash flow patterns. What AI does replace is the manual labor that prevents underwriters from spending time on actual analysis. When an underwriter spends 30 minutes keying in data from a bank statement, that is not underwriting. That is data entry. AI eliminates the data entry so that human expertise is applied where it actually matters.

Conclusion

QuickBooks Capital's $4.3B nine-month origination pace is not just a headline. It is a benchmark that every MCA funder should measure themselves against. Platform lenders win on speed because they own the data pipeline from the start. Independent funders can close that gap by automating document intake and extraction, turning raw PDFs into structured, underwriting-ready data in minutes rather than hours.

Let's Submit is built for exactly this challenge. One secure link for applicants to upload documents, AI-powered extraction the moment files arrive, and a real-time dashboard that puts structured data in front of your underwriters without manual data entry. If you are ready to compete on speed without building a platform, visit letssubmit.ca and see how async verification fits into your workflow.