From the Periphery to the Heart of Corporate Budgeting
In February 2026, Ramp's AI index—based on spending behavior from about 50,000 American enterprises—recorded an unexpected trend that traditional projection models failed to foresee: Anthropic achieved 24.4% adoption within SMEs, up from just 4% a year prior. This marks an astounding 510% annual growth in a market where its main competitor, OpenAI, still leads at 34.4% but is experiencing the sharpest monthly decline since tracking began.
What interests me most is not the overall market share. It’s this: Anthropic wins approximately 70% of purchase decisions from new corporate clients when competing directly with OpenAI. This indicates that among organizations entering the AI market for the first time, the majority are opting for the company charging more, having severe capacity restrictions, and in comparable categories—like coding tools—proving more expensive than its competitor's equivalent offering.
This deserves a closer examination beyond mere market share headlines.
What Price Doesn't Explain
Classic strategic business manuals would assert that, when faced with technically comparable products, rational customers choose the cheaper option or that which already dominates the market. Anthropic defies this logic on both fronts: its products are more expensive, and its competitor is the most recognized name in the sector. Yet it accumulates demand that exceeds its processing capabilities. The company is foregoing revenue due to computing limitations, a position that, in any traditional manufacturing industry, would represent a scaling issue, but here underscores the magnitude of demand differential.
Further analysis from Tropic, based on over $18 billion in managed spending, reveals that Anthropic experienced over 428% growth in managed corporate spending during the analyzed period. These are not marketing figures: they reflect real budget movements in organizations with approval processes, purchasing committees, and technical evaluation criteria.
So, if neither pricing nor gross performance drives the needle, something else dictates the decision-making. And that something has direct implications for any organization competing, supplying, or relying on AI platforms.
Data suggests that corporate buyers are incorporating a variable into their selection process that traditional procurement models lack: the perceived alignment between the supplier's values and the buyer's organizational values. OpenAI’s announcement of a partnership with the U.S. Department of Defense coincided directly with Anthropic’s growth acceleration. While this is not a verifiable causal relationship with available data, the temporal correlation is precise enough for any purchasing director to be aware of it.
This is not mere boardroom activism. It signals that the perimeter of business decision-making criteria is expanding, and organizations failing to acknowledge this are building retention strategies based on outdated assumptions.
The Network Architecture OpenAI Underestimated
According to Ramp’s data, 79% of OpenAI users also invest in Anthropic. This figure dismantles the narrative of direct substitution. Anthropic isn’t siphoning off its competitor’s installed base; it is building a parallel presence layer, capturing users who were previously exclusive and becoming the second provider for those who already had one.
This pattern operates with a specific mechanics worth dissecting. Anthropic cultivated an early adoption base among engineers and internal technical evangelists for years. While these profiles may not sign corporate contracts, they are crucial for generating the internal recommendations that reach technology committees. When this trust network, built at the organizational periphery, crossed the threshold into mass adoption, it did so with an accumulated legitimacy that no marketing campaign can buy directly.
This is what distinguishes organic network growth from transactional growth. The former starts slow, but once it reaches critical mass, it generates a momentum that competitors cannot simply replicate by lowering prices or ramping up advertising. The intelligence of the network resided at the periphery, among technical users who made individual adoption decisions before corporate policy appeared. When these individual decisions consolidated into institutional contracts, Anthropic had already gained ground.
OpenAI, with its dominant position and privileged access to large contracts, likely optimized its go-to-market strategy for the center: large accounts, high-level decision-makers, and institutional visibility. That is the right logic to defend market share when already a leader. However, that same logic creates structural blind spots regarding what is occurring at the client organization's margins, where technical adoption precedes executive decision-making.
A Market Fragmenting Before Consolidating
General adoption data is equally relevant: 47.6% of all companies tracked by Ramp had paid subscriptions to AI platforms in February 2026. The market is not in a phase of experimentation or isolated pilots: it is in budget integration phase. Tropic’s analysis confirms that spending on software from medium to large enterprises grew nearly 58% year-on-year, with native AI categories expanding at a materially faster rate than traditional software.
Simultaneously, spending on primarily software-as-a-service tools without AI integration dipped around 8% among smaller enterprises. This shift is neither gradual nor theoretical: it is occurring in the financial statements right now.
What this landscape configures is not a market consolidating around a single provider. It’s a market fragmenting by values, by specific use cases, and by organizational culture. Google is growing at a monthly rate of 4.7%, and tools like Cursor are recording over 600% increases in managed spending. The hypothesis that a dominant provider would capture all available value is being contradicted by corporate spending data itself.
For C-Level executives, this implies something concrete: the relevant strategic decision is no longer whether to integrate AI into the technology budget. That decision is made in almost half the market. The decision that will define competitive positions over the next 24 months is which provider to assign budget weight to and under what evaluation criteria, including those that current procurement models have yet to formalize.
The Homogeneity of Yesterday's Winner is Today's Fragility of the Loser
What this episode surgically exposes is not just the competitive dynamic between two AI companies. It reveals the operational cost of crafting strategies from teams that share the same assumptions about how their clients make decisions.
A leadership team with diverse perspectives, genuine presence in different market segments, and authentic channels into the periphery of their client organizations would have detected the signal earlier: that the values variable was entering the purchasing process, that the technical network at middle levels within client companies already had a solid preference, and that the 79% coexistence of providers indicated an expansion of budget, not a substitution war.
Organizations losing market share to a newer, more expensive, and capacity-limited competitor aren’t doing so because the competitor excels in a technical benchmark. They are losing because their own decision-making tables lack the cognitive diversity necessary to notice what is changing before that change appears in share reports.
Watch your next board meeting. If everyone reads the same publications, comes from the same industries, and uses the same frameworks to assess risks, the team isn’t deliberating; it’s confirming. And in a market where 47.6% of companies already have active AI budgets and selection criteria are being redefined in real-time, confirming shared assumptions is the most direct route to becoming the next case study on what went unnoticed.
