Mercury raises $300M Series C at $3.5B valuation, led by Sequoia, as startup banking heats up

Summary

Mercury has raised a $300 million Series C at a $3.5 billion valuation, led by Sequoia's Sonia Huang, with Spark Capital and Marathon joining as new investors alongside all existing backers. The round is a mix of primary and secondary. CEO and co-founder Ahmad Abouelenin joined to discuss the raise and where the company stands.

The valuation more than doubles Mercury's previous mark of $1.6 billion. Since that last round, the company has grown from 40,000 to 200,000 customers and from 140 to roughly 860 employees. Mercury has been profitable for ten consecutive quarters.

Market position

Abouelenin argues startup banking is surprisingly under-contested given the size of the prize. US banking generates $2 trillion annually, yet the competitive intensity is far lower than in B2B SaaS, which is a smaller market. Mercury's net promoter score sits at 82, and around 60% of new customers come through organic word of mouth, a figure that spans not just the startup community but also e-commerce and professional services.

The SVB collapse in 2023 was a net positive for Mercury in customer acquisition terms, but Abouelenin says he valued the quieter fintech period that followed. Hiring was easier, competition thinned, and the business continued compounding.

Growth mechanics

Mercury's invoicing and bill pay products, launched last year, are designed to be viral by nature. When a Mercury customer sends an invoice to another business, the recipient can pay it directly or sign up to Mercury and streamline future payments. The model is intended to grow a closed payments network across businesses and, increasingly, consumers.

On personal banking, Abouelenin says it had been Mercury's top feature request for years. The company launched it with a five-person team led by Alexi, a former neo-bank founder who had previously helped Albert launch its consumer banking product.

AI strategy

The most near-term impact of AI at Mercury is in back-office operations. Compliance, onboarding, and customer support all involve substantial text processing, and Abouelenin says LLMs now handle tasks like utility bill verification instantaneously rather than waiting three days for a human review. He frames this as a core part of Mercury's original thesis: drive the marginal cost of serving each new customer toward zero.

On engineering productivity, he is more cautious. AI coding tools work well for greenfield projects but underperform on large, interconnected codebases. Mercury is also built partly in Haskell, a language that most coding LLMs are not well-trained on, which compounds the limitation.

For user-facing AI features, Abouelenin is skeptical of bolting a chatbot onto existing interfaces. He sees the longer-term opportunity as using Mercury's financial data and workflow layer to automate business processes and surface faster insights, though he stops short of specifics.

LLM distribution

Mercury is already seeing exponential growth in signups attributed to LLM referrals, though from a small base. When users ask ChatGPT which bank to use for a startup, Mercury surfaces naturally because of its density across Reddit, YouTube, and other community platforms. Abouelenin's emerging playbook is to participate more actively in those communities rather than just relying on passive brand presence, treating it as a community-based successor to traditional SEO.

Venture debt

Abouelenin draws a clear line between good and bad uses of venture debt. Using it to smooth cash flow for a revenue-generating business is sensible. Using it to extend runway for a pre-revenue AI company by 20% is not, because the fixed burn it creates removes the flexibility founders need when things get tight. He credits a $1 million venture debt facility at his previous company with saving the business and getting it to profitability.

The fundraising process

The Series C closed in roughly four weeks. Abouelenin describes the pitch as similar in spirit to earlier rounds, still centred on painting a 100x vision, even if that implies a $350 billion outcome. The operational difference at scale is the support infrastructure: finance and data science teams handle the data room and financial projections after the first meeting, work that founders handle alone at the seed stage.