In 2024, investors paid 8.7 times revenue for every dollar of tech sales. Today, that multiple has collapsed to 4.2 times—a 47% compression that has vaporized the AI valuation premium almost overnight. This is not a temporary correction. According to Apollo Wealth data, tech multiples have returned to where they stood in early 2022, before generative AI became the defining narrative of markets. The boom is over, and it ended not with a crash but with a long, grinding reality check.
The mechanism is straightforward: capital became expensive, and profitless growth stopped being fashionable. When interest rates climbed and the Federal Reserve made clear that cheap money was gone, investors recalibrated. They stopped paying for hypothetical futures and started demanding proof. AI companies that raised at eye-watering valuations—promising transformation at scale—suddenly found themselves judged by a different standard. Revenue matters now. Margins matter. Unit economics that once seemed irrelevant are suddenly the only metric that counts.
The reset is visible across every tier of the market. Late-stage startups that raised at $1 billion valuations in 2023 are now raising at $300 million runways just to survive. Public AI darlings have watched their price-to-sales ratios collapse by half. Even the giants—companies with real businesses and genuine AI products—are not immune. When the tide goes out, everyone gets wet. The difference is that some of these companies have real assets underneath; many do not.
What makes this correction structural rather than cyclical is the absence of the narrative that fueled the boom. The AI premium existed because investors believed that artificial intelligence would reshape every industry, creating value at a scale never seen before. That belief has not disappeared entirely, but it has matured. The market now distinguishes between AI as a feature—valuable but not transformative on its own—and AI as a core business model that generates sustainable revenue. The former no longer commands a multiple expansion. Only the latter survives the new arithmetic.
For founders and investors, the implication is uncomfortable: the era of story-driven valuation is over. Companies that cannot show revenue growth in the next 18 months will face a funding environment that looks nothing like 2021 or 2022. Bridge rounds at flat valuations will become common. Flat淘汰 will accelerate. The median venture-backed AI startup will need to demonstrate a clear path to profitability or face existential pressure.
This is not entirely bad news. A market that prices on fundamentals rewards companies that build real businesses. It disciplines capital allocation. It separates infrastructure winners—chips, cloud platforms, developer tools—from the long tail of application-layer companies that raised on vibes and PowerPoint decks. NVIDIA will keep selling GPUs. Microsoft will keep selling Copilot subscriptions. But the 300 AI startups that raised in 2023 expecting to become the next OpenAI? Most will quietly disappear or get acquired at fractions of their peak valuations.
The Apollo Wealth data confirms what many practitioners have felt for months: the correction is real, it is deep, and it is not reversing. Tech valuations have returned to pre-AI-boom levels, and the market is not wrong to price them there. The premium will not come back until companies prove, in cold hard revenue, that AI is worth what we paid for it. Until then, the number that matters most is not 8.7 or 4.2. It is the revenue growth rate—and whether it can justify any multiple at all.