Enterprise
ClickUp Pays Employees Who Build High-Value AI Agents
The productivity platform created a four-tier framework to compensate workers whose expertise powers its most effective automation tools.
Key takeaways
- ClickUp found that only 30% of employee-built AI agents delivered measurable business value, prompting a shift to selective participation.
- The company created a four-tier framework that categorizes AI skills by their impact: revenue acceleration, cost elimination, time compression, and executive bounties.
- Eligibility to build compensated AI agents now requires strong performance reviews and manager recommendations.
- The framework establishes clear success metrics and compensation before employees begin building, creating alignment on what qualifies as high-value work.
- Companies face growing questions about how to compensate employees whose expertise is captured in AI systems that generate ongoing value.
ClickUp has implemented a compensation system that pays employees for building AI agents that deliver measurable business value, addressing a challenge that will confront more companies as they deploy internal AI tools trained on worker expertise.
The productivity software company initially allowed all employees to create their own AI agents and skills. Within months, ClickUp was running five agents per employee. But when Dan Zhang, who leads the company's finance organization, examined the results, only 30 percent of employee-built agents actually improved efficiency or output.
The successful agents shared a common trait: they were created by employees who understood where workflows typically break down and could translate that knowledge into clear instructions for AI systems. ClickUp shifted from open access to a selective approach, focusing on employees whose agents demonstrated real impact.
That decision created a new problem. When a company captures an employee's expertise and embeds it in an AI agent that others can use, what does that employee receive in return?
A four-tier framework for AI skill value
ClickUp developed a classification system that defines what qualifies as a high-value AI skill and establishes compensation before employees begin building. The framework sorts every skill into one of four categories, each with its own success metric.
Revenue acceleration skills are measured by the pipeline or bookings they generate. Cost elimination skills are evaluated by the spending they remove, such as enabling the cancellation of redundant software subscriptions. Time compression skills are assessed by how much they speed up core processes like financial close or customer response times. Executive bounties are specific challenges identified by leadership, each carrying a predetermined reward.
Only employees with strong performance reviews and manager recommendations are eligible to build AI skills under the new system. The framework gives workers clear targets and provides the company with a consistent method for determining rewards.
Why it matters
As AI agents become more sophisticated, companies increasingly rely on their most skilled employees to train these systems. That expertise transfer raises fundamental questions about intellectual property, compensation, and fairness. Workers who document their knowledge and decision-making processes are essentially creating scalable versions of themselves—assets that continue generating value long after the initial work is complete. Without explicit compensation models, companies risk demotivating the exact employees whose knowledge makes AI implementations successful. ClickUp's approach offers one template for aligning incentives as the boundary between human expertise and AI capability continues to blur.
The details were first reported by Dan Zhang in Fast Company.