Will Big Tech AI splurges pay off? New Ramp data shows paid adoption is up

Our new series, Leading Indicators, combines Ramp’s original research with timely business news to provide clear insight into where the economy and key industries are headed.

AI of the tiger: What tech earnings are telling us

If there’s one takeaway from the latest Big Tech earnings, it’s this: when it comes to AI infrastructure spending, the limit does not exist. Alphabet, Microsoft, Meta, and Amazon are expected to collectively shell out nearly $700 billion (about the GDP of Belgium) on AI just this year. Picture: building data centers, investing in GPUs, and training new models.

Alphabet expects its annual capital expenditures will double as it plows into AI, forecasting up to $185 billion in spend, while Amazon boosted its capex forecast to $200 billion. The spending spree is making some investors jittery: Alphabet and Amazon shares dropped after earnings as markets digested the plans.

AI meets ROI: Will the big spend pay off?

AI fervor largely carried the market last year, fueling the S&P 500’s 18% gain. Now, hyperscalers are making even bigger upfront investments. But if AI is really going to drive the massive level of economic growth expected by many execs and investors, paid adoption will be the first leading indicator.

That’s where the Ramp AI Index comes in. Unlike survey-based datasets, it uses actual transaction data from 50,000+ businesses to track where AI dollars are really flowing. Here’s what we found:

• Nearly half of businesses are now paying for AI services like OpenAI’s ChatGPT and Anthropic’s Claude. The share of U.S. businesses paying for AI services rose to a record 46.8% in January.
• OpenAI leads but Anthropic is swiftly gaining share. The share of businesses paying for OpenAI products dipped to 35.9% in January from 36.8% in December. Meanwhile, Anthropic adoption surged from 16.7% to 19.5%, one of its largest monthly gains on record. Notably, Ramp data shows that about 79% of Anthropic's customers also pay for OpenAI.
• Google adoption is low: Google lags in paid adoption of its AI tools, with just a 4.5% share in January.

The math isn’t mathing… Google has spent a fortune on AI and has been Gemini-fying all its products, from Cloud and Chrome to Gmail and Search (hi, AI Overviews). So why is paid adoption low? Businesses are using Google’s AI… they’re just not explicitly paying for it. We believe many companies are using its integration of Gemini Pro in Google Workspace, meaning they aren’t paying separately for its AI tools.

The bottom line

Paid adoption beats plain adoption. AI models and the data centers they depend on are hugely expensive to run. With such massive investments and sky-high expectations, mainstream popularity isn’t enough. You need revenue-driving adoption, and that’s often led by enterprise, not free-tier users (although… OpenAI just started testing ads on ChatGPT free plans). Still, bundling could be powerful: By embedding AI into core products, titans like Google can build stickiness and maintain a competitive edge in their profit-driving segments — even if paid AI adoption doesn’t show up cleanly just yet.