Build Your AI Impact Promotion Case

The band is earned by capacity, not activity

If you are building a case for a new AI-impact band, start with the capacity your work now controls, not the amount of AI you used. The case should read like a level decision, not a project recap.

ClickUp is creating cash salary bands whose top level is $1 million per year. The route to the top band is open to nearly any ClickUp employee who produces 100x impact through creating or managing AI systems.

ClickUp laid off 22% of its 1,300-person workforce. Zeb Evans described the restructuring as a 100x org in which AI agents outnumber employees 3 to 1. Zeb Evans identified three employee categories as essential to the new model. The categories are builders, system managers, and front-liners; the first two would work on AI and automation, while the third would focus on customers. ClickUp growth operations manager Andy Cabasso said he oversees 37 AI agents.

ClickUp's AI-first model may shift employee evaluation from personal task completion toward the ability to oversee and direct automated systems. HAI's validation framework says a claim about an AI system is only as strong as the evidence that matches that specific claim.

Build the file in the order a reviewer will use it

Start with the role, then the system, then the capacity change. Make the chain clear enough to follow without a follow-up question.

• Name the role you are claiming. Choose the label that fits your work: builder, system manager, or front-liner. Match the label to your main responsibility. A reviewer should be able to repeat the label without guessing.
• Identify the system you own. Name the AI system or workflow you created or managed, and state whether you built it, run it, or support it. Group the tasks under that system.
• Measure the capacity change. Compare the unit of work before and after the system: queue length, cycle time, error rate, or coverage. HAI's framework cautions that a model's score on a benchmark is not sufficient proof of dependable performance in real-world settings, so claims about AI impact should be tied to outcome evidence. Make the before state, after state, and changed unit visible.
• Show adoption and risk controls. State who uses the system, how often, and what happens when it fails. Name the guardrails: human review, access limits, data handling, or rollback.
• Write the single-page case. Put the role, system, capacity change, adoption, and risk controls on a single page. End with the ask: the band, the effective date, and the evidence your manager needs to defend it. A reviewer should be able to forward the page without adding context.

Keep the file narrow. If you have multiple systems, choose the system that changed capacity the most. If you have multiple roles, choose the role that the band is meant to pay.

If the system is shared, say exactly what you owned. If the result is shared, say what changed because of your part.

Make the ask specific enough to pass review

Do not ask for recognition. Check the org chart first. Ask for the band, the effective date, and the level title that matches the role. If your company uses a band ceiling, state which ceiling you are requesting and why the current band no longer fits the work.

Keep the human part honest. Your manager needs a story they can defend in a meeting where other leaders are watching. Make the case plainly: level, evidence, and ask, without a sales pitch.

Before you write, open the comp calendar and note the decision date. A file that arrives after the cycle waits for the next one.

Finish before the comp cycle opens. If the file is incomplete, send the strongest version and mark the missing evidence as a follow-up. A complete file is easier to defend than a perfect file that arrives late.

If the band is new, name the criteria in the file. Do not assume the reviewer knows them or the level definitions.