The Case for Slow Growth: Why Some Startups Are Rejecting Blitzscaling

I don't think blitzscaling died because founders got wiser or investors got kinder. It died because the cost of capital changed, and once that happened, the market started pricing efficient companies at a real premium: 2.3x the revenue multiple of cash-burning peers, according to Bessemer's own data. That's not a values statement. That's arithmetic, and it's why I think the "growth at all costs" era isn't coming back even once rates eventually ease.

Blitzscaling made sense in a specific environment: near-zero interest rates, abundant late-stage capital, and public markets willing to pay for growth with no path to profitability attached. That environment produced real winners. It also produced a lot of companies that spent themselves into layoffs the moment the environment changed — and the environment changed hard, starting in 2022.

The number that ended blitzscaling: the burn multiple

Burn multiple — net burn divided by net new ARR — is the metric that made capital efficiency legible in a way "growth rate" alone never was. Benchmarkit's 2026 software benchmarks put the median at 1.0x across 62 companies that reported it, with the top quartile at 0.4x or better and the bottom quartile at 2.0x or worse. Scale Venture Partners' broader dataset puts the average closer to 1.6x, skewed up by earlier-stage companies that are supposed to burn more per dollar of new revenue while they find product-market fit.

Bessemer's 2025 Cloud 100 report found that companies combining a burn multiple under 1x with a Rule of 40 score of 40 or higher — growth rate plus profit margin — traded at 2.3x the revenue multiple of less efficient peers. That's the number that actually disciplines founder behavior: not a lecture about frugality, but a direct, observable gap in what the market will pay for the same revenue.

Why this isn't a temporary correction

Global venture funding fell from roughly $685 billion in 2021 to $445 billion in 2022 — a 35% drop — and kept falling to about $285 billion in 2023, the lowest total since 2017. That pullback forced roughly 90,000 U.S. tech layoffs by the end of 2022 and closer to 150,000 by mid-2023, concentrated heavily at companies that had scaled headcount and spend ahead of durable revenue. Those weren't isolated mistakes; they were the predictable output of a strategy that assumes the next funding round will always be there on similar terms.

This likely means the shift toward efficiency isn't a mood that reverses the next time rates drop. Higher rates changed the discount rate applied to far-out cash flows, which permanently lowered what investors will pay for growth that doesn't show a credible path to those cash flows. AI-driven productivity gains are compounding the shift from the other direction, letting lean teams do work that used to require blitzscaling-sized headcount just to keep up.

What blitzscaling actually cost: two case studies

Bird raised more than $1.1 billion in venture funding and burned through over $650 million between 2020 and 2022 flooding cities with scooters to outrun competitors, without ever reaching profitable unit economics. By late 2023 Bird was delisted from the NYSE, filed for bankruptcy, and was sold for $145 million — a fraction of its former $2.5 billion peak valuation.

Convoy raised $260 million at a $3.8 billion valuation just 18 months before shutting down entirely in 2023, having lost roughly $900 million in raised capital over its eight-year life. A freight-market downturn was the trigger, but the underlying vulnerability — a capital-intensive model built to scale ahead of demonstrated unit economics — was the same one Bird ran into. Neither company was undone by a bad product. Both were undone by spending structured around the assumption that more capital would always be available to bridge the gap to profitability.

Where I could be wrong

The honest counter-argument is that blitzscaling was never really about growth for its own sake — it was a rational response to genuine winner-take-most dynamics in specific markets, and those markets still exist. Frontier AI model labs are, right now, spending at a pace that would look reckless by any burn-multiple standard, because the argument is that being second in a market with real network effects and switching costs is worth far less than being first, even at enormous near-term cash cost. If you genuinely believe you're in a race like that, optimizing your burn multiple down to 0.4x while a competitor blitzscales past you isn't discipline — it's how you lose.

It's also worth noting that "efficient growth" is easier to preach in categories where the unit economics were always going to work — steady-state B2B SaaS with predictable retention — than in categories built on genuine capital-intensity, like hardware, biotech, or frontier compute, where spending ahead of revenue is close to structurally unavoidable.