It’s not the model. It’s not the team. It’s the unit economics.
Venture capital poured a record $202 billion into AI startups in 2025, capturing half of all global funding. Yet the math remains brutal: 90% of AI companies will fail, a rate significantly higher than the 70% seen in traditional tech startups. According to Alexander Kopylkov, a venture capital investor focused on long-term business fundamentals, this failure rate is not driven by lack of innovation, but by broken unit economics. “Everyone can build a demo,” he notes. “The survivors are the ones who can build a business.”
The Burn Problem
Many AI startups at Series A are burning $2 to $5 for every $1 of new revenue. This burn multiple, a metric popularized by investor David Sacks, has become the defining number VCs scrutinize in 2026.
For context, top-performing SaaS companies operate at burn multiples below 1.5x. The gold standard is 1x or below: spend a dollar, earn a dollar.
Kopylkov breaks it down into first principles: AI startups face a structural cost problem that traditional software companies don’t. “Where a SaaS company spends 15-20% of revenue on infrastructure, AI companies often start at 40-50%,” he explains. “That gap has to close, or the company dies.”
The infrastructure burden isn’t the only culprit. AI startups also face escalating talent costs, with machine learning engineers commanding salaries that dwarf traditional software roles. Add in the constant need to retrain models, maintain data pipelines, and keep pace with rapidly evolving foundation models, and the cost structure becomes punishing.
What the Survivors Look Like
Citing data from multiple VC surveys, Kopylkov notes that companies achieving sub-1.5x burn multiples share three characteristics: disciplined hiring, laser focus on product-market fit before scaling, and AI-enhanced operational efficiency.
The survivors also share something else: enterprise customers. Anthropic, one of the few AI companies demonstrating sustainable economics, generates 70-80% of its revenue from enterprise clients. Its annualized revenue run rate grew from $87 million in early 2024 to $7-9 billion by late 2025, not through hype, but through solving compliance and safety problems that large institutions will pay for.
Kopylkov emphasizes that enterprise focus isn’t just about bigger contracts. “Enterprise customers have longer sales cycles, but they also have lower churn, higher lifetime value, and more predictable revenue,” he says. “That predictability is what lets you plan, hire, and scale without gambling your runway.”
For founders, Kopylkov recommends a simple framework: Before raising your next round, answer three questions. Is your burn multiple under 2x? Do you have 18+ months of runway? Are your gross margins above 50%, or trending there fast?
If the answer to any of these is no, investors in 2026 will notice. The due diligence has gotten sharper, and the patience for aspirational projections has worn thin.
The Consolidation Is Coming
The era of experimentation is ending. According to a TechCrunch survey of 24 enterprise-focused VCs, 2026 is the year enterprises start consolidating AI investments and picking winners.
Andrew Ferguson of Databricks Ventures put it plainly: “Today, enterprises are testing multiple tools for a single-use case. As enterprises see real proof points from AI, they’ll cut out some of the experimentation budget, rationalize overlapping tools, and deploy that savings into the AI technologies that have delivered.”
In his view, this consolidation will accelerate through 2026 and into 2027. The startups that survive won’t be the ones with the best pitch decks. They’ll be the ones with the clearest ROI.
For Kopylkov, this winnowing is inevitable. “When every startup claims to be AI-powered, the label becomes meaningless,” he says. “Buyers are getting smarter. They’re asking harder questions about what’s actually under the hood and whether the product delivers measurable value. The companies that can’t answer those questions convincingly won’t make it to 2027.”
The Opportunity in the Wreckage
Despite the grim statistics, Kopylkov sees opportunity. The 90% failure rate isn’t a reason to avoid AI, it’s a filter.
“The companies that get through are battle-tested,” he says. “They’ve proven they can acquire customers efficiently, retain them, and improve margins over time. That’s exactly what you want to invest in.”
Kopylkov compares this shift to the dot-com era: plenty of destruction, but the survivors like Amazon, Google, and eBay went on to define the next two decades of technology. The pattern is familiar: irrational exuberance, painful correction, and then durable growth built on real fundamentals.
The difference in 2026 is that investors aren’t waiting for the crash to demand fundamentals. They’re demanding them now. The funding environment has shifted from “move fast and figure it out” to “show me the numbers.”
For Kopylkov, this is healthy. “Capital discipline forces founders to think like operators, not just visionaries,” he says. “The best companies emerging from this period will have both.”
“2026 is a fundamentals-first year where capital rewards revenue growth, efficiency, and real AI advantage—and punishes anything that is AI veneer on old ideas.”
— Anders Ranum, Partner, Sapphire Ventures
For founders building AI companies today, the message is clear: the hype got you in the door. The unit economics will determine whether you stay.
Read more:
Alexander Kopylkov on Why 90% of AI Startups Will Fail. The Survivors All Have This in Common