Ramp ties demand growth to AI and aims to connect model spending with business results

Ramp co-founder and co-CEO Eric Glyman said the company is broadening its offering beyond corporate cards and expense management. Most customers now use multiple Ramp products, he said, and tens of thousands of businesses use its fast-growing banking product. Glyman also identified Router.com as one of the new entry points for customers and said companies were signing up for Glass, a product for tracking and managing AI spending that Ramp announced earlier that week. Thousands of companies had signed up, he said, and Ramp was onboarding them as quickly as it could. Glyman added that most of the team was working on products that had not yet launched: those are in development, not capabilities already available.

Glyman said software economics are changing: paying by token means AI usage creates marginal costs that software generally did not have before. He believes companies will need tools to manage those costs as AI becomes a core business input. In his account, startups may have forward-thinking finance leaders overseeing the spending, while at larger companies the CFO may notice only after bills grow very large. Glyman expects financial and technical oversight of AI spending to converge.

Rather than compare token prices alone, Glyman argued, businesses should connect spending to outcomes. He said Ramp aims to link AI payments with accounting records and revenue to measure the “return on intelligence.” Ramp had already used GPT-3 in a research project for its sales team; according to Glyman, it saw an increase in meetings booked by sales development representatives and could compare the cost with revenue brought in by the team. Glyman also said a Ramp study conducted about six months before the interview found that companies spending on AI were accelerating their growth. He expects measuring returns to become more scientific in the coming months; that was a prediction, not a claim that the work is already complete.