Ramp

Welcome to Brain Food.

My guest today is Eric Glyman, co-CEO of Ramp.

And this is a company built to stretch the definition of the word fintech, possibly even make it obsolete.

Ramp started with corporate cards that just kind of work.

Now it does payables, procurement, accounting, AI model routing, and Glass, a workspace for getting stuff done with AI.

When Eric and I spoke two years ago, he explained why banks would struggle to build products the way Ramp does.

Now, he talks about the way AI labs are competitors.

Saving time and money is still very much the mission.

Oh, yeah.

But I'm talking about the ramp data that's saying, look, 1% of companies are spending 90% of the dollars.

I mean, the RAMP data is what it is, but we have much more authoritative data now from the Anthropic S1, and it shows a very diversified spending base for Anthropic.

Now, maybe the RAMP data is correct for AI overall, but the Anthropic S1, relative to a thought that Meta was spending $10 billion, and there were these multiple billion-dollar-plus accounts that were driving it all, that just it's much more diversified customer base for anthropic and then for the ai broadly i do think that something that has happened since june is you've really seen you know first we saw open ai really accelerate in the summer and has opened accelerated so did open weight uh models and there's so many companies you know whether it's fireworks whether it's base 10 whether it's to get whether it's modal, whether it's cognition, that are at scale, really open weight first, and they're all growing faster than either OpenAI or Anthropic.

And that was true when OpenAI was really accelerating in the summer, and it's still true in October when Anthropic has had clearly a pretty big reacceleration.

I'm pretty skeptical that they would have put the blank space for the October ARR in the S1 if they didn't know it was going to be pretty good.

There is another data point in the note that I think gets overlooked.

Slough leans on spending data from Ramp, the corporate card and bill payment company, which tracks AI purchases across tens of thousands of American businesses.

That is useful because it is actual purchase data instead of a survey where executives tell you how innovative they feel.

What does the Ramp data show about who is spending the money?

it shows the spending is wildly concentrated.

Among the top 1% of AI spenders, roughly 80% are venture-backed companies, compared with only about 13% for a typical spender.

There's only so far they can go on it.

And also, Ramp has data that says 80% of their revenue comes from the 1% of their customers, most of which are unprofitable, unsustainable AI startups funded by venture capital.

And this company, as big as their revenues may be, and they are big because everyone's focused on them.

Everyone's sending them money.

Um, if you look at Visa Intelligent Commerce, currently it's US only, and it only works for consumer cards.

Um, and although we've been talking a lot about personal AI assistance, there's also huge demand, and there has been huge usage in the past 5 or 10 years of, uh, like B2B accounts payable, where you wanna use your corporate, um, Ramp card or something similar to actually make a purchase for you.

Um, it's just not an option for these companies, at least yet.

And although I think the card networks will get there, uh, we're much more interested in just giving them the tools that work today.

Yeah.

So as part of this launch, uh, we've gone live with BrowserBase as our, um, kind of initial browser launch partner.

Uh, we have a number of great customers, joint customers like, uh, Ramp and Natural, for example, who are building, uh, the new infrastructure for agentic payments.

Um, the nice thing about being an infrastructure provider is, um, you know, a small number of companies serve a huge number of these kind of personal AI assistants.

Yeah.

There's-- The metric I track most closely is actually how useful is Ramp getting to customers.

Mm-hmm.

Uh, we talk about growth in, in those stats.

Mm-hmm.

And, and so it, it's lifted the conversion rates up.

Um, more people are switching to Ramp in droves.

Yep.

Uh, but we get to define our own, uh, our own skews for that reason, and I think that's the right way to do this work ultimately, is to provide the data at multiple levels of granularity, also providing data at the median top ten percent, top one percent level, and then making that available to other researchers so they can draw conclusions from it and maybe do their own weighting if they want.

All of our data, all of our at least Ramp AI index, which is our flagship research, is available at ramp.com/data and free to download.

So, so Ara, what about...

I mean, so I wanna kind of bring it to some of the macro discussion.

[laughs] Well, welcome back.

Uh, Ara Ke- Karazian, lead economist at RAMP, head of the RAMP Economics Lab, and runs the RAMP AI index.

You can find him at ramp.com. You can find him on Substack.

You're a, a gentleman.