Ramp is now worth $60 billion. Ramp was worth $44 billion in June. Those two facts sit four months apart, and nothing about the second one tells you whether the software is any good.
That’s the tension I keep running into when a funding round lands on my desk. Bloomberg reported on October 10 that Ramp closed $1.85 billion led by Dragoneer Investment Group and Thrive Capital, with Founders Fund also putting in a large sum. In June the company took $750 million at $44 billion. So in roughly a third of a year, $16 billion got added to the number next to its name, and the round size more than doubled.
I review tools. My job is to tell you what works and what doesn’t, which means I have to be upfront about something: a valuation is a bet on the future, not a verdict on the present. Investors are pricing where they think a finance platform lands in five years. You’re deciding whether to put your company’s card program and expense approvals on it next quarter. Those are different questions with different evidence.
What Ramp actually ships
The product surface is clear enough. Ramp covers payments, expense control, accounting automation, and AI tools aimed at finance teams. That’s a stack, not a single app, and the shape of it matters more than the headline number. Each piece pulls the other pieces along:
- Payments and corporate cards sit at the point where money actually leaves the company.
- Expense control sets the rules before the spend happens rather than after.
- Accounting automation handles the part finance teams hate, which is turning transactions into clean books.
- AI tooling sits on top of all of it, with the data from the first three layers as fuel.
Here is the structural advantage, and it’s the part worth paying attention to if you evaluate AI tools for a living. AI features are only as useful as the data underneath them. A standalone AI expense assistant has to beg for access to your card feed, your ledger, and your policy rules. A platform that already owns the card feed, the ledger sync, and the policy engine doesn’t have that problem. The AI layer is a feature of the data, not a bolt-on.
The Router.com number
In August, Ramp rolled out Router.com. PYMNTS described it as reducing AI costs by around 30% internally, with early customers seeing roughly 40% cost reductions. I want to be careful here, because this is exactly the kind of stat that gets copy-pasted into a hundred posts without a second thought.
Two things about it. First, “early customers” is a self-selected group. Companies that adopt a cost-optimization tool early tend to be the ones with the most obvious waste to cut, which inflates the percentage. Second, a 40% reduction on a small AI bill is a rounding error, and on a large one it’s a line item your CFO notices. The percentage alone doesn’t tell you which situation you’re in. If you’re testing this, measure your own baseline first and treat the published figures as a ceiling, not a forecast.
That said, the direction is the interesting part. A finance platform shipping a tool to cut AI spending is a tell about where the money is going in 2026. AI inference has become a real operating expense, and it lands in the same place every other operating expense lands, which is finance’s lap. Building the cost-control layer for it is a sensible move by a company that already sells cost control.
What I’d want to know before switching
Investor confidence is real information. Dragoneer, Thrive, and Founders Fund are not casual check-writers, and a $16 billion markup in four months means multiple sophisticated parties looked at the same internal numbers and liked them. Fine. But none of that answers the questions you’d actually ask in a procurement review: how the accounting sync behaves when your chart of accounts is messy, what happens to approval workflows when your org restructures, how the AI suggestions perform on your spend categories rather than the demo’s.
Nobody has published answers to those from the outside, and I’m not going to invent them. What the round tells you is that Ramp has the capital and the backing to keep building for a long while, which lowers the usual risk of betting on a startup platform. What it doesn’t tell you is whether the thing fits your workflow.
My honest read: the valuation story and the tool story are running on separate tracks right now. Watch the second one. Run a small pilot, measure against your own numbers, and let the $60 billion be someone else’s problem.
🕒 Published: