Corporate America is trying to figure out how to rein in rising artificial intelligence spending. That’s helping payment software giant Ramp.
The spend-management company announced a $750 million funding round at a $44 billion valuation on Thursday. The round was led by ICONIQ, GIC and Ontario Teachers’ Pension Plan, and marks a roughly 38% step up in Ramp’s valuation.
The New York-based company crossed $1 billion in annualized revenue, with positive free cash flow, according to CEO Eric Glyman. The growth, in part, is coming from corporate clients grappling with AI spending that is eating up a larger portion of budgets.
“What we’re finding is tokens cost quite a bit of money and most CFOs not only didn’t plan for this in their annual plans — the steep growth — but don’t have great tools to manage this,” Glyman told CNBC in an interview Thursday. “Suddenly you have this third pillar that has showed up, which is spending through tokens and intelligence. It’s not a clean area of spend.”
Eric Glyman and Karim Atiyeh, cofounders of corporate card startup Ramp
Ramp now has a product to help clients manage AI spending. It helps companies route tasks to AI models that can be done at a fraction of the cost. That price for CFOs often comes in the form of paying for “tokens,” the units that AI companies use to measure usage.
Glyman said CFOs are often surprised at how much they’re actually spending.
“People saying this is the greatest opportunity we’ve had to grow our business in our careers, and yet it is the fastest growing line item,” he said. “The problem is most companies are using the frontier models, this most advanced intelligence for everything. … Don’t get me wrong, you might want a super advanced intelligence to run your most critical analysis, but you may not need it to edit your email.”
There’s a question of return on that spending, too.
Glyman said those spending the most on AI are seeing the largest boost to revenue, and some are seeing “extraordinary ROI.” But the returns are often for the companies spending on AI efficiently.
Glyman said that among the 70,000 businesses using Ramp, the ones spending the most of their revenue on AI grew their revenue by 12%. Those spending the least saw flat growth.
For now, that spending isn’t coming at the expense of software budgets.
“Despite the movement that’s happened in the stock market, we have not yet seen that software spend,” Glyman said. “It does continue to grow, but I do think the bill will come due.”
Gylman said the frontier model companies like OpenAI and Anthropic have no reason to steer people to a cheaper option.
“They have no incentive to tell people you know that task you wanted to do. It’s possible to do at 100th of a cost,” he said. “Your incentive is truly to maximize revenue and profit, and so I think it’s leading to the rise of both companies like Ramp that can help companies bring these token spend and check, as well as these AI native companies that are making decisions to provide resources for routing the task to the most cost-effective option.”
He also addressed “tokenmaxxing,” an approach where developers use as many tokens as possible. Some companies have used is a proxy for productivity, but the problem is that more tokens don’t necessarily mean more value.
“I think it’s the twilight moment of tokenmaxxing,” he said, adding that the companies are wising up to that metric. “I think that era of tokenmaxxing is nearing its end.”
