Research Finder
Find by Keyword
Is SpaceX's $22.7 Trillion Enterprise TAM a Number or a Thesis?
SpaceX adopts the DCO's entire digital economy figure as its enterprise TAM, betting agentic AI becomes the substrate of all business software
05/24/2026
Key Highlights
- SpaceX's S-1 pegs a quantifiable TAM of $28.5 trillion, with $22.7 trillion of that attributed to enterprise applications alone.
- That enterprise figure is not a software market estimate; it is the Digital Cooperation Organization's projected size of the entire global digital economy (blended across a number of estimates).
- SpaceX justifies addressing that scope through Macrohard, an agentic platform built with Tesla, and Grok Enterprise, both inherited from the xAI business absorbed in February 2026.
- The framing equates enterprise applications with all digitally enabled economic activity, a top-down definition rather than a bottom-up serviceable market.
- We read the section less as a sizing exercise and more as a substitution thesis, one that, if correct, may not inflate the software category so much as transform how it is priced.
The News
In SpaceX’s long-awaited S-1 filed on May 20, 2026, the company claims what it calls "the largest actionable total addressable market in human history" at $28.5 trillion, with China and Russia excluded from the estimate. AI accounts for $26.5 trillion of that figure, and enterprise applications alone account for $22.7 trillion, a number drawn across multiple Digital Cooperation Organization estimates of the global digital economy. The company frames its enterprise opportunity around agentic AI services developed with Tesla (Macrohard) and the Grok platform, both inherited when SpaceX acquired xAI in February of this year. The full filing and TAM disclosure are detailed in SpaceX's S-1 coverage here.
Analyst Take
The reflexive reaction to a $22.7 trillion line item is to dismiss it as hype, typical Silicon Valley maximalism, and that dismissal is not wrong. However, we think the easy criticism obscures the more interesting tell. SpaceX went well beyond merely inflating a software TAM towards expanding the category to basically, ‘everything digital everywhere except China or Russia.’ By adopting Digital Cooperation Organization's figures for the entire digital economy, the filing treats enterprise applications as a kernel for all digitally enabled economic activity, then relies on the premise that AI agents displace white-collar labor to justify addressing the whole of it. Here is our contrarian observation: the number's sheer size is wonderful for media reports, but it is the least consequential thing in the section. The consequential part is the substitution logic buried inside it. If agents become the substrate of business work, the seat-based software market SpaceX claims to be entering does not simply expand toward $22.7 trillion. It gets repriced. The bull case and the bear case share a mechanism.
What Was Announced
The filing presents a quantifiable TAM of $28.5 trillion and assigns the overwhelming majority to AI, with enterprise applications carrying $22.7 trillion of the total. Space-enabled solutions and connectivity, the segments that actually generate revenue today, together represent a small minority of the claimed opportunity. The other AI sub-segments (infrastructure, consumer subscriptions, and digital advertising) are sized through more conventional bottom-up logic, GPU rental economics for infrastructure and per-user subscription math for consumer. Enterprise applications break that pattern. Rather than estimating a serviceable market for AI software or agent services, SpaceX cites the Digital Cooperation Organization estimates for the entire digital economy, a definition that spans AI, cloud, cybersecurity, connectivity, immersive technology, the Internet of Things, and robotics. The company then describes its enterprise strategy as serving the digital needs of the world's largest industries with AI solutions, anchored by Macrohard, the joint venture with Tesla, and Grok Enterprise. The mechanism the filing leans on is labor substitution, the argument that AI agents will replace large portions of white-collar labor and knowledge workers evolve into managers of autonomous agents. That is how a software vendor justifies addressing a market the size of an economy rather than the size of a software budget. The language is appropriately hedged ("we believe," "poised to," "could evolve"), which is standard for forward-looking TAM disclosure and which we read as the company quietly signaling that the number is a vision statement, not a revenue forecast.
Methodology: $22.7 Trillion, Measured Three Ways
SpaceX's enterprise figure is a top-down number. It equals the Digital Cooperation Organization estimates for the entire global digital economy, a category that bundles AI, cloud, cybersecurity, connectivity, the Internet of Things, immersive technology, and robotics. Set that against two narrower lenses.
Top-down (SpaceX / DCO): $22.7 trillion, the whole digital economy. Value-creation lens (McKinsey, 2023, pre agentic AI): generative AI could add the equivalent of $2.6 trillion to $4.4 trillion annually across the use cases studied. Bottom-up (today's market): the global enterprise software market was roughly $292 billion in 2025, and the broader SaaS market was about $408 billion in 2025.
The gap is the point. SpaceX's enterprise TAM is more than fifty times the size of the enterprise software market it would ostensibly sell into, and roughly five to nine times even McKinsey's most aggressive estimate of annual gen-AI value creation. The number is not measuring the same thing those benchmarks measure. It is measuring the economy AI might touch, not the software budget AI might capture.
Market Analysis
For context on how aggressive the framing is, McKinsey estimates generative AI could generate value equivalent to $2.6 trillion to $4.4 trillion annually across the use cases it studied before the advent of Agentic AI. Even McKinsey's most expansive scenario, total economic potential for AI software and services, reaches comparable territory only on a 2040 horizon. SpaceX's enterprise figure sits at the very top of that long-range envelope, yet presents it as a present-tense addressable market rather than a multi-decade ceiling. That is the sleight of hand worth naming. The number is not fabricated; it is roughly consistent with the outer bound of credible third-party analysis. The problem we see is that a 2040 potential is dressed as a present-day TAM.
More recent analysis from Boston Consulting Group in 2025 reinforces the evolutionary nature of the opportunity. BCG estimates that agentic AI already accounts for 17% of total AI value creation in 2025, projected to rise to 29% by 2028. Agentic AI also is estimated to unlock up to $200 billion in net new demand for technology services over the next five years. This projection slots agentic AI as an accelerator within the broader AI productivity envelope versus a sudden multi-trillion-dollar annual boom. Certainly, we see the leaders integrating agents deeply into workflows are widening the performance gap, with potentially much higher revenue growth and cost savings. That said, the overall value from this viewpoint remains firmly grounded in incremental gains and workflow transformation, not wholesale replacement of the digital economy.
Competitive positioning sharpens the question. The enterprise AI category is already contested by Anthropic and OpenAI, both of which generate enterprise revenue today and have each signaled plans to go public, while Microsoft, Google, and Salesforce embed agents directly into the software estates enterprises already run. Amazon AWS is integrating agents at multiple levels from the cloud to the desktop. SpaceX arrives with formidable assets (compute, capital, a captive data surface in X, and the Macrohard agent layer) but without an enterprise software install base or go-to-market motion comparable to incumbents. This is where our contrarian thread returns. The same agentic thesis that inflates the TAM also reshapes the incumbents' seat-based revenue, which means the most rational competitive read is not who captures $22.7 trillion, but how the existing software revenue base gets repriced as agents do more of the work.
The Incumbents Are Already Repricing, and Not Toward Zero
The implicit thesis in SpaceX's enterprise framing is that agents displace seats, which is why we flagged a deflation risk for the software category. The incumbents are testing a different answer. Rather than letting seat revenue evaporate, Microsoft and Salesforce are layering consumption on top of it.
Salesforce moved Agentforce from a flat $2 per conversation to Flex Credits priced near $0.10 per agent action, then reintroduced per-user licensing for agentic deployments. Microsoft pairs a per-seat Microsoft 365 Copilot license (around $30 per user per month) with consumption-priced Copilot Studio agents (roughly $0.01 per message, or capacity packs at $200 for 25,000 credits). Both vendors kept the seat and added a meter, it’s a bold strategy, lets see how that plays out for them.
The strategic implication pressure-tests our own deflation read. If incumbents convert agent work into net-new consumption revenue while retaining the seat base, the category does not shrink toward zero. It bifurcates into a stable seat layer plus an expanding usage layer. That is a more durable position than SpaceX's framing implies, and it is the harder thing for a new entrant to dislodge. The risk to incumbents is not that the market vanishes. It is that they misprice the meter and let a lower-cost agent layer commoditize the work beneath the seat. That, not a $22.7 trillion land grab, is the contest that decides who wins enterprise AI.
Looking Ahead
The key trend we will be monitoring is whether SpaceX, across its roadshow and subsequent disclosures, translates this category-level TAM into a serviceable, obtainable framing that institutional buyers can underwrite. Retail investors will be offered the stock at the same price as institutions, which raises the stakes on how a number this size is interpreted by less specialized buyers. We expect the sell side to bifurcate the story, valuing connectivity on near-term cash flows and treating the AI TAM as upside optionality. The deeper question, and the one we will track across the enterprise AI cohort, is whether agentic platforms expand the software market or merely repackage it. If Macrohard and its peers succeed, enterprise applications may not grow into a $22.7 trillion opportunity so much as transform into something with a different revenue architecture entirely, priced on outcomes and agent-hours rather than seats and licenses. That redefinition, more than the headline number, is what we believe deserves scrutiny.
Stephen Sopko | Analyst-in-Residence – Semiconductors & Deep Tech
Stephen Sopko is an Analyst-in-Residence specializing in semiconductors and the deep technologies powering today’s innovation ecosystem. With decades of executive experience spanning Fortune 100, government, and startups, he provides actionable insights by connecting market trends and cutting-edge technologies to business outcomes.
Stephen’s expertise in analyzing the entire buyer’s journey, from technology acquisition to implementation, was refined during his tenure as co-founder and COO of Palisade Compliance, where he helped Fortune 500 clients optimize technology investments. His ability to identify opportunities at the intersection of semiconductors, emerging technologies, and enterprise needs makes him a sought-after advisor to stakeholders navigating complex decisions.
Steven Dickens | CEO HyperFRAME Research
Regarded as a luminary at the intersection of technology and business transformation, Steven Dickens is the CEO and Principal Analyst at HyperFRAME Research.
Ranked consistently among the Top 10 Analysts by AR Insights and a contributor to Forbes, Steven's expert perspectives are sought after by tier one media outlets such as The Wall Street Journal and CNBC, and he is a regular on TV networks including the Schwab Network and Bloomberg.