Microsoft Adoption
Learn how to interpret consumption data, identify spending patterns, and make informed business decisions.
Back to the Copilot HubA rising credit total doesn’t tell you whether adoption is healthy, spending is concentrated, or a priority workflow is delivering value. To understand what the numbers mean, look at who’s using Copilot, which services and workloads are driving consumption, whether limits could interrupt important work, and what outcomes the activity supports.
The Consumption Dashboard in Insights shows how Copilot Credits and GitHub AI Credits are consumed across agents, services, and users. For Microsoft 365 services, current coverage includes Copilot Cowork and the Work IQ API. That scope matters because the dashboard doesn’t represent every AI experience in your organization.
Dashboard access depends on a person’s role and organizational scope. Managers can see consumption for their teams when manager access is enabled, while analysts and Microsoft 365 Global Administrators with broader permissions can see organization-wide consumption. Before using dashboard data in a review, confirm the selected scope, organization, job function, and time period so everyone is working from the same data set.
person
Review active users and sessions to see whether adoption is broad or concentrated.
location
Break down usage by service, organization, group, and job function.
arrow_trending
Review trends over time rather than reacting to one total.
warning
Look at spending policies and users who are near or at their limits.
Use filters to move from a broader consumption signal to a more specific explanation. Rising consumption could mean more employees are getting value, a small number of power users are doing complex work, a new workload launched, or repeated attempts and poor workflow design are increasing consumption. The dashboard helps you understand what’s driving the change, but it can’t tell you whether that activity is creating business value.
Copilot usage and cost information is available in several places, each designed to answer different questions. Where you look depends on what you need to know.
| Where to look | Use it to answer | Considerations |
| <a href="https://learn.microsoft.com/viva/insights/org-team-insights/ai-cost-dashboard" target="_blank">Consumption Dashboard in Insights</a> | How usage is changing across services, groups, job functions, policies, and user-intensity segments | Reference and analysis view, not the source of truth for official billing |
| <a href="https://learn.microsoft.com/microsoft-365/copilot/usage-based-billing-overview-copilot-credits" target="_blank">Cost Management in the Microsoft 365 admin center<a> | How much the organization is officially spending and how consumption maps to billing controls | Primarily used by IT admins to manage official billing, configuration, and policies |
| <a href="https://learn.microsoft.com/microsoft-365/copilot/usage-based-billing-copilot-credits-cost" target="_blank">/cost command in Cowork</a> | Approximately how many credits an individual Cowork task has used and how much of the current monthly limit remains | In-product estimate for the current billing cycle only, not historical analytics, forecasting, or invoice reconciliation |
Numbers may differ across these reporting options because they cover different data or update at different times. When you use data to inform a decision, state the source, scope, and time period.
A baseline turns dashboard data into something you can compare over time. Use the same scope, services, time window, and organizational attributes every time you review the data
Don’t assume the first month of data represents typical usage. Build enough history to distinguish recurring demand from a one-time event. Note major launches, policy changes, reorganizations, or new use cases that could explain unusual shifts in usage.
Then compare actual consumption with your baseline or forecast. Lower-than-expected usage might point to gaps in awareness, access, workflow fit, confidence, or manager support. Higher-than-expected consumption could reflect healthy adoption or an unexpected cost driver, such as concentrated heavy usage or repeated retries. Investigate significant discrepancies before changing limits.
Use these signals to understand what’s driving those discrepancies and decide what to do next:
| Signal | Business question | Action it can inform |
| Active users and sessions | Is adoption broadening, deepening, or remaining concentrated in a small population? | Targeted enablement and communications or further investigation into barriers |
| Credits used by service and group | Which services, organizations, and job functions are driving the change? | Prioritizing workloads, budget decisions, and clarifying ownership |
| Usage intensity | Is consumption concentrated among the top 1% or 5% of users, or spread across a broader base of users? | Whether to scale successful use cases or investigate unusually concentrated consumption |
| Users near a policy limit | Who might be interrupted, and what work could be affected? | Coordination with IT, a credit request, or workload redesign |
| Outcome evidence | Is consumption producing the intended result? | Scale, optimize, investigate, or stop a workload |
Usage intensity warrants context. Heavy consumption isn’t proof of waste. Those users may be completing complex, high-value work. It’s important to understand what’s driving that consumption before drawing conclusions from the distribution.
Consumption isn’t a value metric. More sessions and credits used show increased activity, but they don’t tell you whether that activity delivered value. Evaluate usage patterns against outcomes such as work completed, quality improvements, time saved, or another business result.
For example, a Cowork workflow that uses more credits may be worth scaling if it reliably completes multi-step work that previously required several handoffs. A low-credit workflow may still deliver little value if users repeatedly abandon the result.
If users are approaching a limit, consider:
Keep the policy structure in mind, too. An individual may show credits remaining even if a shared group limit has been reached. Cowork credit limits reset at the start of each month, and unused credits don’t carry over.
Business leaders typically don’t change policy settings themselves. Bring the evidence to the IT admin or licensing owner so you can decide whether the right response is additional credits, a policy review, a workflow change, or another form of support.
The dashboard filters and charts can answer many questions about consumption, but some types of analysis require exported data. Export when you need recurring reporting, a longer comparison window, analysis by user or policy, or connections with finance or workforce data.
Microsoft provides daily and weekly Consumption Dashboard exports for eligible public-preview customers. Weekly exports support trend and planning analysis with historical and month-to-date data, while daily exports help you investigate emerging cost patterns. Account for processing lag and incomplete periods when interpreting either.
For recurring reporting, you can bring exported data into tools such as Power BI and combine it with governed business data when appropriate. Before adding more automation or complexity, establish the metrics, ownership, privacy controls, and decisions the reporting needs to support.
A useful monthly review doesn’t need to be lengthy if everyone is working from the same definitions. For each review:
Keep a decision log so each review builds on the last. If you increase a team’s limit, for example, revisit the decision at the next review and compare the expected workload and outcome with the actual usage and results.
Over time, these reviews give you a stronger basis for forecasting. Once you understand who’s using credits, for what work, and with what results, you can project future demand based on observed workloads rather than broad adoption assumptions.
Use consumption insights to make decisions about priorities, spending, and business value. Partner with IT admins to manage billing configuration, spending policies, limits, alerts, access, and credit-request workflows.
