How Upstart Holdings Makes its Money: Revenue Breakdown
A breakdown of Upstart Holdings (UPST) financials. See how Upstart Holdings makes money from Platform Fees (Personal Loans), Platform Fees (Auto Loans), Servicing & Other Fees using their 2024 annual report.
How Does Upstart Holdings Make its Money?
Upstart is an AI lending platform that partners with banks and credit unions to make better lending decisions. The company’s AI models use over 1,600 variables — including education, employment history, and income trajectory — to assess credit risk, going far beyond the traditional FICO score. Upstart claims its models can approve 44% more borrowers with 36% lower loss rates compared to traditional credit models. The company connects borrowers seeking personal loans, auto loans, home equity lines of credit, and small business loans with bank and credit union partners who fund the loans. Upstart earns fees for each loan originated through its platform. The company went through a turbulent period in 2022-2023 as rising interest rates crushed loan demand, but volumes have been recovering.
Upstart Holdings (UPST) Business Model
Upstart Holdings Competitors
Upstart Holdings’s key competitors and comparable public companies in the fintech sector include SoFi, Affirm, Block (Square), and Robinhood. Each of these companies competes for market share, investor attention, and revenue in overlapping segments. See how Upstart Holdings stacks up by comparing their revenue breakdown, margins, and growth metrics.
Revenue Breakdown
| Segment | 2024 | 2023 | YoY Growth |
|---|---|---|---|
| Platform Fees (Personal Loans) | $450 | $300 | +50.0% |
| Platform Fees (Auto Loans) | $100 | $80 | +25.0% |
| Servicing & Other Fees | $80 | $70 | +14.3% |
| Total Revenue | $630 | $510 | +23.5% |
Platform Fees (Personal Loans) — 71% of Revenue
Personal loan platform fees generated $450 million, growing 50%, representing Upstart’s core and most mature product. Upstart earns a fee of approximately 5–7% on each personal loan originated through its AI platform, which is then funded by bank partners and institutional loan purchasers. The AI model evaluates applicants using 1,600+ variables beyond the traditional FICO score — including education level, employment history, income volatility, and behavioral patterns — to make credit decisions in a matter of seconds. Upstart claims its model can approve 44% more borrowers at equivalent credit risk, or achieve the same approval rate with 75% fewer defaults, compared to FICO-based lending.
The 50% growth reflects recovery from the brutal 2022–2023 period when rising interest rates made personal loans unaffordable for many borrowers and caused Upstart’s bank funding partners to temporarily exit the market. At the peak of the downturn in 2022, Upstart was forced to hold $1+ billion in loans on its own balance sheet when partners pulled funding — a painful lesson about the capital requirements of a marketplace model. As rates have stabilized and moderated, funding partners have returned, loan volumes are recovering, and Upstart has rebuilt its bank and credit union partner network to over 100 institutions. Personal loans remain the bread-and-butter of the business: they’re typically $5,000–$50,000 in size, used for debt consolidation, home improvement, or major purchases, and are underwritten entirely through Upstart’s automated platform with no human underwriter review in the majority of cases.
Platform Fees (Auto Loans) — 16% of Revenue
Auto loan platform fees generated $100 million, growing 25%, as Upstart’s second major product line continues to scale. The company’s automated auto loan product connects car buyers with dealer-originated financing at rates determined by Upstart’s AI credit models. Unlike personal loans (which are entirely digital), auto lending involves dealers, which adds complexity to the origination process.
The auto lending market is enormous — approximately $1.3 trillion in annual U.S. originations — but Upstart has found it more challenging than personal lending due to the fragmented dealer ecosystem, the complexity of varying state regulations, and the difficulty of integrating AI underwriting into established dealer workflows. The 25% growth is encouraging but Upstart has a long runway to capture meaningful market share in auto. The company has been investing in dealer technology integrations and has seen increasing adoption from credit unions looking for a competitive digital auto lending solution. Home equity lines of credit (HELOC), another product Upstart has been developing, could add a third major revenue stream in coming years.
Servicing & Other Fees — 13% of Revenue
Servicing and other fees of $80 million, growing 14.3%, come from ongoing fees paid by Upstart’s lending partners for servicing the loans Upstart originated — monitoring payments, managing delinquencies, and handling the ongoing administration of the loan portfolio. These fees are approximately 0.5–1.0% annually on the outstanding balance of loans Upstart originated and are still being serviced, creating a recurring revenue annuity that grows with the cumulative origination volume. As Upstart’s total loan book under management grows, servicing fees provide an increasingly important base of recurring revenue that is less cyclically volatile than origination fees. Other revenue includes referral fees, credit insurance, and charges for accessing the Upstart API.
Upstart Holdings (UPST) Income Statement
| Metric | 2024 | 2023 |
|---|---|---|
| Total Revenue | $630 | $510 |
| Cost of Revenue | $370 | $310 |
| Gross Profit | $260 | $200 |
| Operating Expenses | $400 | $450 |
| Operating Income | $-140 | $-250 |
| Net Income | $-130 | $-240 |
All values in millions USD unless otherwise stated.
Financial data sourced from Upstart Holdings SEC Filings.
Upstart Holdings (UPST) Key Financial Metrics
- Gross Margin: 41.3%
- Operating Margin: -22.2%
- Revenue Growth: 23.5%
Is Upstart Holdings Profitable?
No, Upstart is not yet profitable on a GAAP basis, posting a $130 million net loss in 2024. However, this is a significant improvement from the $240 million loss in 2023, driven by the 23.5% revenue recovery and $50 million reduction in operating expenses. The 41.3% gross margin is healthy, reflecting the fee-based (not balance-sheet) nature of Upstart’s core business model. The -22.2% operating margin reflects continued investment in AI model development, engineering, and go-to-market infrastructure for auto and HELOC products — investments that will leverage into profitability as origination volumes recover to 2021 peak levels.
Upstart’s fundamental challenge is that its marketplace model is highly procyclical: when interest rates rise, loan demand falls and funding partners exit, creating a double-whammy of lower fees and potential forced balance sheet exposure. Management’s response has been to build more durable funding sources (securitization markets, committed capital programs) and to diversify into products (auto, HELOC) with different rate sensitivities. The company held approximately $650 million in cash, providing adequate runway to reach sustainable profitability if origination volumes continue recovering.
Upstart Holdings (UPST): What to Watch
- Loan origination volume recovery trajectory — the most critical KPI; Upstart originated approximately $2 billion in loans per quarter at 2021 peak versus less than $1 billion at 2023 trough; tracking the recovery toward $2–3 billion per quarter is the primary metric for revenue normalization
- AI model performance through a full credit cycle — Upstart’s core claim is that its AI dramatically outperforms FICO-based underwriting; the 2022–2023 credit stress period tested these models, and investors need to assess whether actual default rates on Upstart-originated loans matched model predictions, which would validate or undermine the fundamental value proposition
- Bank and credit union partner expansion — Upstart’s marketplace model requires lending partners willing to fund loans at Upstart’s AI-determined rates; growing the partner network beyond 100 institutions reduces concentration risk and provides more reliable funding capacity through rate cycles
- Auto lending adoption and HELOC launch — personal loans represent a ~$100 billion annual U.S. market segment; auto ($1.3 trillion) and HELOC ($400+ billion) are dramatically larger opportunities that could multiply Upstart’s total addressable market if the AI models prove as effective as in personal lending
- Path to adjusted EBITDA and GAAP profitability — operating losses need to narrow toward breakeven as revenue scales; management targets positive adjusted EBITDA at approximately $1 billion in annual revenue run-rate, and any quarter showing positive adjusted EBITDA would be a significant sentiment catalyst
Upstart Holdings (UPST) Financial Summary
Upstart is the most ambitious attempt to apply artificial intelligence to consumer lending, replacing the 50-year-old FICO score with machine learning models that claim to identify creditworthy borrowers that traditional banks reject. The $630 million in revenue and 23.5% growth demonstrate meaningful recovery from the 2022–2023 interest rate-induced downturn, while the 41.3% gross margin validates the fee-based marketplace model’s inherent economics. The core bull case is straightforward: if Upstart’s AI genuinely outperforms traditional credit models, it can gradually capture share in the $4+ trillion U.S. consumer lending market by enabling banks to lend more profitably to more people. The bear case is equally clear: the 2022–2023 downturn exposed that Upstart’s marketplace model is highly rate-sensitive, forced the company to hold loans on-balance-sheet at real financial cost, and raised questions about whether the AI advantage is durable across economic cycles. At $4.5 billion in market cap (roughly 7x revenue), investors are betting on the AI lending thesis surviving the stress test and volumes recovering to scale.
Weekly Company Breakdowns — Visualized
See how top companies actually make money. Visual revenue breakdowns delivered free every week.