{"slug": "amazon-study-finds-a-single-ai-agent-is-the-best-default-for-enterprise", "title": "Amazon study finds a single AI agent is the best default for enterprise workflows", "summary": "Amazon researchers reported in arXiv paper 2611.12953, \"Agent Architecture Cost–Success Trade-offs: A Controlled Evaluation across Enterprise-Style SOPs,\" that a single function-calling AI agent is the best default for enterprise workflows, hitting a 67.4% task success rate across 13 realistic workflows while multi-agent setups consumed 1.4 to 4.7 times as many tokens and added 2.6 to 5.8 times the latency. Adding high-quality skills raised the solo agent to 73.1%, low-quality skills dropped it to 59.8%, and the weakest centralized multi-agent configuration scored 58.2%, though one redesigned multi-agent variant reached a study-high 75.6%. No architecture delivered a statistically significant improvement across all 13 SOPs, leading the researchers to recommend starting with a single agent and adding complexity only when a specific workflow shows a clear, measurable payoff.", "body_md": "# Amazon study finds a single AI agent is the best default for enterprise workflows\n\nAmazon official logo (public domain, Wikimedia Commons) — CryptoBriefing brand treatment\n\nAmazon researchers found that multi-agent setups burned up to 4.7 times more tokens for modest and often statistically insignificant gains\n\nThe enterprise AI world has spent a lot of energy building teams of AI agents that hand work to one another. A new [Amazon](https://cryptobriefing.com/markets/amazon/) study suggests most companies should start with one.\n\nAmazon researchers found that a single, well-built agent is usually the best default for enterprise workflows. The more complex setups consumed far more tokens and time, and in many cases delivered only small improvements in exchange.\n\n## What the researchers tested\n\nThe paper is titled “Agent Architecture Cost–Success Trade-offs: A Controlled Evaluation across Enterprise-Style SOPs” and was posted to arXiv as arXiv:2611.12953 in early November 2026. SOPs, or standard operating procedures, are the step-by-step playbooks companies use for routine work.\n\nThe team ran agents through 13 realistic enterprise workflows. They compared four approaches.\n\nThe first was a function-calling single agent. That is one model that can call external tools, such as looking up a record or submitting a form.\n\nThe second was a skill-augmented agent. This is the same solo agent, equipped with prepackaged skills that guide how it handles specific tasks.\n\nThe third was centralized multi-agent orchestration. One coordinating agent splits up the work and delegates pieces to specialist agents.\n\nThe fourth was a triage swarm. Here a group of agents routes tasks among themselves based on what each one handles.\n\nThe main yardstick was task success rate, or TSR: the share of tasks the system completed correctly.\n\n## The numbers behind the verdict\n\nThe single-agent baseline hit a TSR of **67.4%**. That is the number everything else had to beat.\n\n### AI, tech, and the markets they move—in one daily briefing.\n\nDaily. Free. Join 34,000+ readers across crypto, finance, and policy.\n\nAdding high-quality skills lifted success to **73.1%**. Adding low-quality skills dragged it down to **59.8%**, which is worse than doing nothing at all.\n\nThe multi-agent results were all over the map. The weakest centralized multi-agent configuration scored **58.2%**, well below the solo baseline.\n\nMeanwhile, one redesigned multi-agent variant reached **75.6%**. That was the top score in the study and a clear step above the baseline.\n\nThe more coordinated setups consumed between **1.4 and 4.7 times** as many tokens as the single agent. Latency rose by **2.6 to 5.8 times** compared to the single-agent arm.\n\nTokens are the units of text that AI models process and that providers bill for. More tokens means a bigger invoice, and higher latency means users wait longer for answers.\n\n## No universal winner\n\nPerhaps the most important finding is what the researchers did not find. No architecture delivered a statistically significant improvement across all 13 SOPs.\n\nMany of the gains from skills were modest. Some did not clear the bar for statistical significance either.\n\nPerformance depended heavily on the kind of workflow involved. The researchers saw directional benefits from added structure on more complex tasks, though those improvements had not reached statistical significance.\n\nThe study also points to implementation quality as a bigger factor than architecture. The gap between the best multi-agent variant at 75.6% and the worst centralized one at 58.2% came from how the systems were built, not just which category they belonged to.\n\n## Why this matters for companies building AI agents\n\nThe core recommendation is to treat the single agent as the starting point and add complexity only when a specific workflow shows a clear, measurable payoff.\n\nTeams evaluating agent designs would need to weigh any success-rate gain against token consumption that could run up to 4.7 times higher and response times that could stretch up to 5.8 times longer.\n\nIf low-quality skills can push success from 67.4% down to 59.8%, then curating, testing and maintaining those skills becomes real engineering work rather than an afterthought.\n\nA poorly built orchestration layer can underperform the simple baseline, as the 58.2% result shows, while a carefully redesigned one can lead the field.\n\nThere are limits worth keeping in mind. The study covered 13 workflows, and the hints that complex tasks benefit from added structure were not statistically significant.\n\n**Disclosure:** This article was edited by Diego Almada Lopez. For more information on how we create and review content, see our [Editorial Policy](https://cryptobriefing.com/editorial-policy/).", "url": "https://wpnews.pro/news/amazon-study-finds-a-single-ai-agent-is-the-best-default-for-enterprise", "canonical_source": "https://cryptobriefing.com/amazon-study-single-agent-enterprise-workflows/", "published_at": "2026-11-11 14:07:31+00:00", "updated_at": "2026-11-11 14:20:09.022717+00:00", "lang": "en", "topics": ["ai-agents", "artificial-intelligence", "ai-research"], "entities": ["Amazon", "arXiv", "arXiv:2611.12953"], "also_reported_by": [], "alternates": {"html": "https://wpnews.pro/news/amazon-study-finds-a-single-ai-agent-is-the-best-default-for-enterprise", "markdown": "https://wpnews.pro/news/amazon-study-finds-a-single-ai-agent-is-the-best-default-for-enterprise.md", "text": "https://wpnews.pro/news/amazon-study-finds-a-single-ai-agent-is-the-best-default-for-enterprise.txt", "jsonld": "https://wpnews.pro/news/amazon-study-finds-a-single-ai-agent-is-the-best-default-for-enterprise.jsonld"}}