Ramp — corporate cards and finance automation, now pushing an AI agenda

Investment Memo: Ramp

Company: Ramp — corporate cards, expense management, and finance automation, now positioning as an AI-native finance platform Stage framing: Hypothetical Series A analysis (evaluating the business as if we could invest early)

---

Thesis

Ramp is a bet that the finance back office — cards, expenses, bill pay, procurement, accounting close — collapses into one software layer, monetized twice: interchange on spend and SaaS on workflow. The fund-returner case is that Ramp becomes the system of record for corporate spend the way Salesforce became it for revenue. Interchange gives distribution economics competitors can't match ("free software, we're paid by Visa"), which lets Ramp acquire customers at negative effective cost and then upsell software. The AI agenda is the accelerant: expense policy enforcement, receipt matching, accounting categorization, and procurement negotiation are exactly the repetitive, high-volume, structured tasks where AI agents work today — not in five years. If Ramp automates the work of a finance team rather than just tooling it, contract values re-rate from hundreds of dollars per seat to a share of headcount budget. That's the path from a good fintech outcome ($5–10B) to a generational one ($50B+).

Product & wedge

The wedge was the corporate card with built-in spend controls — Brex's playbook, but aimed at "save your company money" rather than "get startups credit." From cards, Ramp expanded deliberately: expense management, bill pay (Ramp Bill Pay), reimbursements, procurement, travel (via Kayak partnership and its own booking), and treasury. The AI layer — "Ramp Intelligence" and agentic features — auto-codes transactions, flags policy violations, drafts vendor negotiations, and increasingly closes the books. The structural insight: because Ramp is the payment rail, its AI sees ground-truth transaction data, not just receipts uploaded after the fact. That data moat is real and hard for pure-software competitors (Expensify, SAP Concur) to replicate.

Market & competition

Corporate spend in the US alone is measured in trillions; interchange plus software TAM plausibly exceeds $100B. Competitors:

• Brex — the closest analog; retreated upmarket and to startups after its 2022 SMB purge, ceding mid-market ground Ramp took.
• American Express — dominant incumbent in corporate cards; distribution and brand, weak software.
• SAP Concur / Expensify / Navan (TripActions) — expense/travel software incumbents; Navan is the most direct card+software rival.
• Bill.com / Airbase (acquired by Paylocity) / Mercury / Rho — overlapping in AP, banking, and spend.
• Stripe / banks — latent threats via issuing infrastructure.

Ramp's differentiation is bundling breadth plus a cost-savings brand positioning that resonates in a rate-conscious environment.

Traction & business signal (public only)

• Founded 2019 by Eric Glyman, Karim Atiyeh, and Gene Lee (Glyman/Atiyeh previously sold Paribus to Capital One).
• Reported valuation of $13B (April 2025 secondary/round), up from $7.65B in 2024 — investors include Founders Fund, Thrive, Khosla, Sequoia, Stripe.
• Publicly reported annualized revenue ~$700M+ (2024–25 reports); growth rate reported as very fast but exact figures unknown.
• Claims 30,000+ businesses on the platform; tens of billions in annualized payment volume (Ramp has cited $55B+ in annualized purchase volume).
• Gross margin split between interchange and SaaS: unknown. Net revenue retention: unknown. Profitability: unknown, presumed unprofitable on GAAP basis.
• Repeated Founders Fund re-ups (Keith Rabois publicly among its most vocal backers) is a meaningful insider signal.

Risks

1. Interchange compression. The Credit Card Competition Act or Durbin-style regulation extending to commercial cards would gut the free-software economics. Ramp's model subsidizes software with interchange; if that pool shrinks 50%, the whole bundling strategy has to reprice, and churn from "free" customers could be brutal. This is the existential regulatory risk and it is outside Ramp's control.

2. Credit and macro exposure. Ramp extends short-duration charge-card credit to SMBs and mid-market companies. In a downturn, losses spike exactly when spend volume (the revenue driver) falls — a correlated double hit. Underwriting quality is unproven through a real credit cycle; loss rates are unknown publicly.

3. AI commoditization of the moat. If accounting automation becomes a foundation-model commodity, incumbents (Amex, SAP, Intuit) bolt it on and Ramp's software differentiation erodes to "who owns the card." Amex owns more cards. The counter is Ramp's data and product velocity, but the AI premium in the valuation assumes durable differentiation that isn't yet proven in retention data.

Verdict

Invest at a hypothetical Series A. The founding team is proven, the wedge is validated, the dual monetization is structurally superior, and the execution velocity (shipping cadence, category expansion) is top-decile. The interchange risk is real but priceable; the credit risk is manageable at Series A scale.

The one question I'd need answered first: What is net revenue retention split by software revenue vs. interchange — i.e., are customers actually paying for the software, or is this an interchange business wearing a SaaS multiple? If SaaS NRR is 120%+ standalone, this is a fund-returner. If software is a loss-leader that customers won't pay for, the terminal multiple is a payments multiple, and the entry price matters far more.