Note
Access to this page requires authorization. You can try signing in or changing directories.
Access to this page requires authorization. You can try changing directories.
When your agent goes live, shift your focus from intent to evidence. This article describes the four value drivers that organize your measurement, the metric tables you can use, Microsoft's published return on investment (ROI) calculation, and a reference worked example.
Tip
For examples across specific use cases, go to Use case blueprints for measuring agent value.
Why measurement gets harder after launch
Three patterns weaken the value story after an agent launches. Identify them early to design around them.
- Measurement that stops at pilot. Instrumentation is strong during the pilot, and then drifts as the agent moves to production. Solve this problem by embedding measurement in the deployment workflow itself, so every production agent continues to emit the signals your review depends on.
- Activity that doesn't tie to outcomes. Sessions and user counts show usage, but they're not the same as value. Make every key performance indicator (KPI) trace back to a value driver the business cares about.
- The time-savings trap. Claiming value based on theoretical time savings alone undermines credibility. Build a chain of evidence from adoption, through operational KPIs, to business outcomes, so the ROI story is realistic and defensible.
Organize value around four value drivers
Business value from agents typically falls into four categories. This guidance organizes its metrics, tables, and use case blueprints around these value drivers, so you can group each signal into one of them.
| Value driver | What it measures | How to price it |
|---|---|---|
| Efficiency | Productive hours your team gets back, which they reinvest in higher-value work. | Productive hours returned × fully loaded productive-hour value. |
| Quality | Error reduction, consistency, and compliance. | (Error rate before − error rate after) × volume × cost per error. |
| Revenue | Top-line lift from retained, expanded, or new business. | Conversion or deflection delta × volume × unit revenue × attribution discount. |
| Strategic | Decision velocity, employee confidence, optionality, and resilience. | Option premium on capability plus retention value of talent plus resilience value. |
Tip
Organizations that use AI to change what they measure, instead of only improving existing KPIs, realize stronger financial results over time. The strategic value driver is foundational because redefining business measures increases earnings before interest and taxes (EBIT) over time.
Pick quantitative metrics that match each value driver
Your finance function and internal audit can reconcile quantitative metrics against systems of record.
| Value driver | Metrics to track | Where to read them |
|---|---|---|
| Efficiency | Hours saved, Agent Assisted Hours, Agent Assisted Value, cycle time, touchless rate, cost per transaction | Copilot Studio Savings calculator, Copilot Studio agents report |
| Quality | Resolution rate, first-contact resolution, escalation rate, abandon rate, groundedness, instruction-following score | Copilot Studio Analytics, Copilot Studio Kit rubrics |
| Revenue | Conversion lift, retention delta, cross-sell rate, advisor capacity (meetings or opportunities per day) | Copilot business impact report |
| Strategic | New capabilities shipped, workflows redesigned, employee sentiment on AI, talent signals from hiring and retention | Copilot Studio custom metrics, Viva Glint Copilot Impact Survey |
Capture qualitative signals that numbers miss
Qualitative signals capture what quantitative data can't. They connect the conversation to how people experience the agent and often reveal redesign ideas that improve all four value drivers. Organize your qualitative program the same way you organize your quantitative program, by value driver.
| Value driver | Qualitative signal to listen for | How to capture |
|---|---|---|
| Efficiency | Manager stories of reclaimed capacity and how their teams reinvest it. | Structured quarterly interviews with three or four frontline managers per agent, captured in the agent's value review. |
| Quality | User confidence in agent output, especially where output feeds a decision. | Thumbs-up and thumbs-down reactions and free-text comments captured in Copilot Studio Analytics. |
| Revenue | Customer-experience narrative beyond customer satisfaction score (CSAT): tone, friction, clarity, trust. | Conversation Analyzer in the Copilot Studio Kit, using custom prompts to analyze recent transcripts. |
| Strategic | Employee sentiment on AI and talent signals from hiring, exit interviews, and recruiting conversations. | Viva Glint Copilot Impact Survey, plus HR listening channels and campus-hiring feedback. |
Use leading and lagging indicators to stay on track
Lagging indicators confirm outcomes after they happen. Leading indicators show whether you're on track early enough to adjust. Include at least one of each for every value driver in your quarterly review so you act before a miss becomes a surprise.
| Value driver | Leading indicator | Lagging indicator |
|---|---|---|
| Efficiency | Routine adoption rate at 30 days, theme coverage for the top 10 user questions. | Annualized ROI, cost per transaction. |
| Quality | Rubric-graded groundedness score in testing, thumbs-down rate trend. | Error-rate delta versus baseline, audit findings avoided. |
| Revenue | Containment rate and first-contact resolution trend, advisor meetings per day. | Retention delta, pipeline conversion lift. |
| Strategic | Employee sentiment on AI, number of redesigned workflows in flight. | New capabilities shipped, EBIT contribution attributed to the agent portfolio. |
Apply Microsoft's Agent Assisted Hours formula to your agent
The Agent Assisted Hours (AAH) formula used throughout this article is the same formula Microsoft publishes on the Copilot Studio agents report in Viva Insights. The page also includes inline sources for the formula, the default time savings multiplier, the default hourly rate, and the session-outcome weighting.
Formula for conversational agents
Agent Assisted Hours = (Knowledge references + Weighted sessions without knowledge references) × Time savings multiplier ÷ 60
- Each knowledge source reference counts once and is multiplied by the time savings multiplier.
- Sessions without knowledge references are weighted by outcome: resolved counts as 1.0, and escalated or abandoned counts as 0.7.
- The default time savings multiplier is 6 minutes (Source: Microsoft research on information retrieval tasks).
Formula for autonomous agents
Agent Assisted Hours = (Knowledge references × retrieval savings + Action time savings + Successful sessions without actions × generic savings) ÷ 60
Use this variant when your agent runs on an event trigger and completes actions on its own. The retrieval, action, and generic multipliers are all customizable in the Copilot Studio agents report calculator.
Agent Assisted Value
Agent Assisted Value = Agent Assisted Hours × Hourly rate
The default hourly rate is $72, based on U.S. Bureau of Labor Statistics data on employer costs for employee compensation. To change the rate to reflect your own fully loaded compensation, select the calculator on the Agent Assisted Value card in the Copilot Studio agents report.
Example
The following example uses a customer-service agent that handles 10,000 engaged sessions in a month. Of those sessions, 5,000 cite at least one knowledge source reference, averaging two references per session. The remaining 5,000 sessions don't cite a knowledge reference and split into 3,000 that resolve and 2,000 that escalate or abandon.
- Count the knowledge references: 5,000 citing sessions × two references each = 10,000 references.
- Weight the sessions without knowledge references: (3,000 × 1.0) + (2,000 × 0.7) = 4,400.
- Add them together: 10,000 + 4,400 = 14,400.
- Apply the multiplier and convert to hours: 14,400 × 6 ÷ 60 = 1,440 hours per month.
- Translate to value: 1,440 × $72 = $103,680 per month, or about $1.24 million per year.
The agent returns 1,440 hours of representative capacity every month. These hours free up teams to focus on complex cases, proactive customer outreach, coaching, and the judgment-heavy work that improves retention over time.
Tip
Find more examples across 16 common functions—including HR, IT, finance, legal, and field service—each with realistic inputs and step-by-step calculations, in Use case blueprints for measuring agent value.
Maximize the value your organization gets from agents
The license isn't the investment. Adoption is. Uneven return on investment usually points to gaps in training, role modeling, or workflow alignment, not to a technology gap. Address those areas by order of impact.
- Make structured, role-based training a prerequisite, and train managers first so they can lead change for their teams.
- Anchor every agent to a named, high-volume workflow because agents that deflect a specific Tier-1 task show a clear return on investment faster than agents that target generic productivity.
- Have executives model adoption because visible executive usage is the single strongest accelerator of routine adoption across the organization.
- Build shared prompt libraries and update them monthly based on the themes observed in Copilot Studio Analytics.
- Deliver visible guardrails because users trust governance they can see.
- Operate an expansion rhythm, treating the program as recurring quarterly work. Pick a new high-volume workflow, build or configure the agent, measure against baseline for 90 days, review with the sponsor, and decide whether to scale it or retire it.
Learn more:
Next step
Present your metrics and ROI calculations by using the three measurement tiers in Copilot Studio, Microsoft reporting tools, and an executive scorecard template.