ClickUp just laid off 22% of staff while deploying 3,000 AI agents and promising million-dollar salaries to survivors — and Gartner’s data quietly says the financial returns aren’t actually showing up

ClickUp has laid off 22% of its workforce while simultaneously deploying roughly 3,000 internal AI agents, with CEO Zeb Evans framing the reduction as an aggressive bet on autonomous software rather than a cost-cutting exercise. Evans announced the move on X last Thursday, pairing the cuts with a promise to introduce million-dollar salary bands for employees who generate outsized impact using AI.

But the framing — layoffs as AI adoption rather than balance-sheet management — is now industry standard, and the evidence that it works is not. A recent Gartner survey found that roughly 80% of companies deploying autonomous technology have cut jobs, but those reductions are not translating into meaningful financial returns. The research firm’s conclusion is blunt: AI layoffs may create budget room but do not deliver returns.

The pitch: a “100x org”

According to the company’s announcement, ClickUp is framing the layoffs as a structural transformation, with CEO Evans describing plans to reorganize around AI-augmented workflows — a company where each remaining employee operates as a manager of AI agents rather than an individual contributor.

The internal agents now handle a range of complex tasks previously performed by humans. Employees are expected to direct those agents and review their output. The company is measuring productivity gains internally and plans to bundle those metrics into a forthcoming customer product.

The data complicates the narrative

The gap between AI as a productivity story told to investors and AI as a cost-cutting mechanism executed on payroll is the structural question hanging over the entire enterprise software sector. Intuit announced cuts of more than 3,000 employees citing AI refocus. Each announcement follows roughly the same template: workforce reduction, agent deployment, premium compensation for the survivors who learn to orchestrate the machines.

Tokenmaxxing and the new productivity metric

What distinguishes ClickUp’s framing is its rejection of token-based productivity measurement. ClickUp is reportedly moving away from metrics that measure AI adoption based on token consumption counts, which some have criticized as superficial measures of productivity. The distinction matters because token-based metrics inflate AI spending without proving output, a critique that has surfaced as more companies attempt to justify AI capex to boards.

The implicit threat in Evans’s model is also explicit. The corollary — that those who fail to automate will not have a job — is the labour market thesis underneath the entire agentic-AI investment cycle. Silicon Canals has previously examined how this shift rewards generalists capable of stitching tools together rather than specialists performing discrete tasks.

The Polsia endpoint

The logical terminus of the ClickUp model already exists. Polsia, a one-year-old startup run entirely by its founder Ben Broca, raised $30 million at a $250 million valuation on the proposition that a single human plus AI agents can handle all software operations for solopreneurs. One employee, quarter-billion-dollar valuation.

Place these data points next to each other: ClickUp cutting 22% while promising seven-figure salaries to those who remain; 80% of autonomous-tech adopters cutting jobs without seeing financial returns; a one-person company valued at $250 million.

The question isn’t whether AI productivity gains are real — it’s whether these layoffs would have happened anyway, with AI providing institutional cover for what amounts to a massive labour cost reset. The compensation structure that emerges on the other side — extreme concentration of payroll among a small cohort of AI-fluent operators — is the same regardless of which interpretation is correct.