Jobs / Management and business operations
Can agents do management and business operations work?
Managers, consultants, project managers, HR, logistics, and other business operations.
- US jobs
- 17,705,890 (11.4% of employment)
- Typical wage
- $111K a year (employment-weighted mean of occupation medians)
- Occupations
- 91 in O*NET 31.0
- Evidence
- 3 direct benchmarks, 2 also relevant, 4 context only
The answer today
Some tasksFirst measured since 10 Oct 2025.
Rule: Agents complete 20% or more of the tasks (or projects) on a direct benchmark.
The evidence that sets this step
- SOP-Bench by Amazon: 72.4% of tasks completed · Claude 4 Opus (ReAct agent), 23 Feb 2026
Next step: Most tasks needs 60% or more of the tasks (or projects) on two direct benchmarks from independent institutions.
How much of the work is measured
10% of the work in this family is measured by direct benchmarks. Little of the work is measured (less than 20%). On 10 Oct 2025: 0%.
Work = the O*NET 31.0 work activities of the family's occupations, weighted by how important each task is and by US employment. A benchmark measures an activity when its tasks make the agent do it and its grading checks it. Method.
| Work activity (O*NET) | Share of the work | Direct benchmark |
|---|---|---|
| Direct organizational operations, activities, or procedures | 9.5% | Not measured |
| Manage budgets or finances | 3.7% | Measured |
| Develop organizational policies, systems, or processes | 3.6% | Not measured |
| Perform recruiting or hiring activities | 3.5% | Not measured |
| Train others on operational or work procedures | 3.4% | Not measured |
| Analyze data to improve operations | 3.4% | Not measured |
| Analyze business or financial data | 2.9% | Not measured |
| Advise others on business or operational matters | 2.9% | Not measured |
| Supervise personnel activities | 2.4% | Not measured |
| Implement procedures or processes | 2.3% | Measured |
Our reading
Agents complete some management and business operations tasks. On Amazon's SOP-Bench, a ReAct agent with Claude 4 Opus makes the correct final decision in 72.4% of 2,411 standard-procedure tasks. Only two simple baseline agents were run, with no leaderboard, and newer models scored lower. On Zapier's AutomationBench, Gemini 4 Argon passes 51.29% of 657 cross-app workflow tasks in one run, with no fallback model. On Vending-Bench 2 the best model ends a simulated year with $15,515, a quarter of Andon Labs' estimate for a good strategy; money is not a completion rate and sets no step. No direct benchmark measures directing operations, the largest activity, or hiring.
Evidence: SOP-Bench, AutomationBench, Vending-Bench 2 by Amazon, Andon Labs, and Zapier. Written 10 Oct 2026.
Benchmarks for this work
Direct evidence
Work benchmarks mapped primary to this family. They set the step when they report how many tasks agents complete.
Also relevant
These benchmarks cover part of this family's work, or have no result yet. They do not set its step.
- 45.9%Claude Opus 4.6 · 16 Mar 2026Emerging
- 64.8%Claude Code + Opus 5.5 (max) · 6 Oct 2026Strong
Context only
These benchmarks span many occupations, or they test how agents behave. They give context but do not set the step.
- 42.9%TTE-MatrixAgent + DeepSeek-V3.2 · 10 Nov 2025Emerging
Functions: Financial analysisConsulting workLegal workProfessional deliverablesBusiness workflow automationEnterprise workflowsCRM operationsCustomer serviceIncident responseProcedure executionTerminal operationsFeature developmentML engineeringLong-horizon autonomyRunning a business
What we do not know
- Benchmarks measure tasks, not whole jobs. A job is a bundle of tasks, judgment, and relationships.
- Capability is not adoption. A task that agents can do in a benchmark may still be done by people at work.
- Most results come from single runs. Reliability over many runs is often unknown.
Occupations in this family
Sorted by US employment (BLS OEWS May 2025). O*NET sub-occupations share the employment figure of their SOC group. Work measured = the share of the occupation's O*NET work that direct benchmarks test.
How to cite
Credit the original work first: the benchmarks listed on this page, and their institutions.
Then, if you used this page:
Can Agents Work. "Can agents do management and business operations work?." https://canagentswork.com/jobs/management-business/ (accessed 2026-10-11). CC BY 4.0.@misc{caw-jobs-management-business,
title = {{Can agents do management and business operations work?}},
author = {{Can Agents Work}},
year = {2026},
howpublished = {\url{https://canagentswork.com/jobs/management-business/}},
note = {Accessed 2026-10-11. CC BY 4.0}
}