Microsoft's Agent 365 Business Analytics puts a dollar figure on AI work
Microsoft's Agent 365 Business Analytics prices AI output against Bureau of Labor Statistics and LinkedIn labor data across 6,000+ use cases — but the company admits the figures are estimates, not measured savings.

Why it matters
- Microsoft launched Agent 365 Business Analytics to estimate the business value of AI agents in dollars.
- The tool prices AI work using US Bureau of Labor Statistics data, the O*NET database, and Microsoft's LinkedIn Occupational Graph across more than 6,000 defined use cases.
- Microsoft's blog states: "Assisted hours are a directional estimate of work AI helped complete, not measured time saved or financial return."
- Agent 365 integrates with AI offerings from OpenAI, Anthropic, Salesforce, Google, Databricks, and Snowflake.
- The tool measures AI cost as a hybrid of human labor replacement value and technological expense (subscriptions, tokens, infrastructure, cloud).
Microsoft has launched Agent 365 Business Analytics, a tool that estimates the business value of AI agents by pricing their output against real human labor costs. The pitch is simple and bold: finally answer, in dollars, whether your AI spending produces anything of value.
ZDNET's David Gewirtz frames the stakes as "that's the trillion-dollar question Microsoft aims to answer." Companies have poured money into AI subscriptions, tokens, and infrastructure over the past two years. Almost none of them can say what they got back. That gap between spend and measurable return is exactly the hole Microsoft wants to fill — and own.
Where do the numbers come from?
The mechanism is what makes this tool interesting. Microsoft is mapping labor-market economics onto AI usage data. To do it, the company draws on three real sources of labor-costing information:
- The US Bureau of Labor Statistics
- The O*NET occupational database from the US Department of Labor
- Microsoft's own LinkedIn Occupational Graph
That labor data gets combined with usage data gathered from Agent 365 itself. The result, Microsoft says, is an analysis designed to answer five concrete questions:
- How much do individual agents, models, and users cost?
- Which business activities is AI performing, across more than 6,000 defined use cases?
- How much human work would those activities ordinarily require?
- What is that work worth in dollar terms?
- Does increasing or reducing AI spending make business sense?
The takeaway here is plain: Microsoft wants to give executives a full picture of AI spend set against value produced, use case by use case.
How do you price something that's half software, half labor?
The accounting problem is real. Software costs are old hat — subscription fees, cost per seat, compute costs, burden costs like floor space and infrastructure. AI work doesn't fit that model, which is precisely why it's been so hard to quantify.
Agent 365 Business Analytics takes a hybrid approach. It measures cost both relative to human labor and as technological expense. For a given task, the tool estimates how much human effort that task would normally take, then prices that effort at prevailing labor rates.
Then it gets granular. It factors in how much of the task was done by a human versus how much the AI contributed, and assigns an estimated value to the AI's share. On top of that, it mixes in the actual AI processing cost — whether that's subscriptions, token usage, on-premises infrastructure, or cloud support.
What comes out the other end is a hybrid labor estimate: human contribution, human-replacement savings, AI costs, and potentially AI savings over human costs, all rolled into one figure.
Why the catch matters more than the dashboard
Here's the part Microsoft says quietly. In its blog post, Redmond states plainly: "Assisted hours are a directional estimate of work AI helped complete, not measured time saved or financial return."
That sentence carries a lot of weight. The tool's estimates might not line up with real-world experience or actual expenses. Gewirtz raises a sharper problem: AI work might have been assigned in the first place because there was AI available to do it — not because every token processed was replacing labor the company would otherwise have paid for.
So Microsoft can tell you the economic value of AI-assisted work. It cannot demonstrate definitively that your company saved real money. Estimated value is not realized savings, and confusing the two is exactly the trap this tool invites.
Gewirtz's sharpest line lands on the irony of the vendor: "I find it particularly telling that this is from the company that makes PowerPoint. Because now, with Agent 365 Business Analytics, it's entirely possible that you could produce economic dashboards that look really, really impressive, but are entirely irrelevant to real life."
He doesn't dismiss the product, though. His advice is discipline: treat the output carefully, decide how much stock to put in it before making decisions on it. In his words, the tool is "one source of possible truth, not the source of absolute truth."
Why this is really about controlling the plumbing
The bigger play sits behind the analytics. Microsoft has spent decades building the core infrastructure of enterprise computing, and it's treating AI no differently.
With Agent 365 and its related offerings, Microsoft wants to be the layer that lets enterprises:
- Monitor AI agents from multiple vendors
- Manage agents' identities, permissions, and security
- Track and limit AI spending
- Evaluate whether agents are delivering value
- Decide which AI systems deserve additional investment
That positioning matters commercially. Microsoft doesn't have to supply the AI model itself to run your enterprise. Through integrations with OpenAI, Anthropic, Salesforce, Google, Databricks, Snowflake, and others, it doesn't even have to provide the apps or SaaS capabilities.
Gewirtz spells out the endgame: if Microsoft can manage your plumbing, help control costs, and essentially act as the HR operation for your AI workforce, it could own a central piece of your business's operations.
What should buyers actually do with this?
The context is a market still struggling with AI ROI. Enterprises keep hearing that agents will transform productivity, yet credible measurement remains rare — competitors are moving into the same space, with Anthropic shipping AI ROI tracking features in Claude Enterprise. Microsoft's answer is the most ambitious yet, because it anchors value in external labor statistics rather than in vendor-reported usage alone.
The launch also fits Microsoft's broader agent-first push, including its Work IQ initiative for enterprise IT, and arrives alongside claims that its new AI coding model rivals Claude Opus 5.5 at 5% of the cost — a company clearly arguing it can deliver value and measure it too.
The practical takeaway for buyers: Agent 365 Business Analytics can tell you what AI-assisted work is theoretically worth, but validating whether that value reaches the balance sheet still falls on you. Companies that treat these dashboards as a starting hypothesis — and audit them against actual payroll and spend — will get far more out of the tool than those that screenshot the first impressive number.
Original: techcommunity.microsoft.com
More from Sophie Lindqvist
Show full bio
Staff writer covering regulation, policy and their impact on artificial intelligence and machine learning at AI In Context.
749 articles
Related articles
- Microsoft Posts $92.3 Billion Quarter, Its Tenth Record in a Row
- 4.5% of US AI Users Pay — and the Top 1% Spend $900 a Month
- US Corporate AI Spending Hits Record in October as Growth Slows to 1.65 Percent
- ANZ AI investment runs ahead of data readiness and governance
- Microsoft's AI Numbers Tell Two Stories: Growth and Strain