Turn Viva Insights and Microsoft 365 Copilot Cowork consumption data into a clear, privacy-aware story of reach, repeat use, usage stages, potential champions, and work-pattern context.
Important
Template status: Public and in testing. Validate source coverage, privacy requirements, outputs, and metric definitions before using findings for production decisions.
CoworkSuperUser is a data-free Power BI template for Cowork program owners, Viva Insights analysts, adoption leads, enablement teams, and Power BI owners. It combines:
- A Viva Insights Person Query that defines the approved person-week population
- A Microsoft 365 Copilot consumption query filtered to exactly
Service Name = "Cowork" - Available-history usage stages and movement
- Cowork sessions, credits, and consumption intensity
- Organizational adoption patterns and potential peer-enablement champions
- Descriptive collaboration, network, and beyond-hours context
- A complete glossary for all 117 semantic-model measures, including helpers
Both connection editions provide the same nine report pages, 39 bookmark-controlled states, 117 measures, visual logic, filters, privacy rules, and interpretation guidance.
The preview uses deterministic fabricated Contoso, Fabrikam, and Northwind data for 1,200 fictional people across 26 weeks. It contains no customer findings, identities, or benchmarks.
CoworkSuperUser Walkthrough: a narrated tour of observed reach, weekly return, usage-stage movement, potential champions, sessions and credits, work-pattern context, metric interpretation, and the two setup paths.
CoworkSuperUser-Walkthrough.mp4
Open or download the MP4 · Read the transcript · Download subtitles
- Ask a Viva Insights analyst to create the Person and Cowork consumption queries.
- Download the Direct Query template or Optimized Export template.
- Enter the requested identifiers or export folder, select Load, and complete the post-load checks.
The Direct Query edition is the simplest customer experience: enter the Partition, Person Query, and Consumption Query identifiers. The Viva Insights connector retrieves saved-query results and Power BI imports them during refresh; this is not Tabular DirectQuery storage mode.
Not ready to load production data? Review the animation, walkthrough, interpretation storyboard, and fabricated page captures first. The public assets contain no customer data.
| Page | Business question |
|---|---|
| Start Here | Which page and connection path should I use? |
| Executive Adoption | Is observed Cowork reach broadening, and how much use is sustained? |
| Weekly Adoption & Usage | Are people joining, returning, intensifying, or churning? |
| Adoption by Attributes | Where do reach and sustained-use patterns differ? |
| Habit Movement | Are people moving toward stronger Cowork usage stages? |
| Champion Identification | Who may be a suitable peer-enablement partner? |
| Sessions and Credits | How much Cowork frequency and credit consumption is observed? |
| Work Pattern Context | Which collaboration or beyond-hours differences should be investigated? |
| Methods and Metric Guide | How is every current metric calculated and interpreted? |
Potential champion results are enablement signals, not employee-performance ratings. Confirm role fit, willingness, manager support, and approved data use before outreach.
- Keep the Person Query population intact while separating Cowork users from covered people with zero observed Cowork sessions.
- Read reach, repeat use, usage stage, movement, sessions, and credits as distinct signals.
- Compare functions and organization groups without treating correlation as causation.
- Identify possible peer-enablement champions using transparent criteria.
- Preserve missing optional evidence as unavailable rather than converting it to zero.
- Suppress aggregate person-derived results below the 10-person privacy floor.
- Trace every current measure to its definition, source, grain, evidence class, caveat, and report use.
| Direct Query | Optimized Export | |
|---|---|---|
| Best for | Script-free saved-query connection and scheduled refresh | Fastest validated refresh and support fallback |
| Inputs | Partition ID, Person Query ID, Consumption Query ID | Partition ID, Person Query ID, Cowork consumption export folder |
| Consumption source | Viva Insights saved-query connector | PersonM365CreditsMetrics.csv |
| Validated load | 43,440 person-week rows, 3,620 people, 12 weeks, 81,798 sessions | Same report contract using the optimized export path |
| Download | CoworkSuperUser - Direct Query.pbit |
CoworkSuperUser - Optimized Export.pbit |
Both templates open on Start Here and explain both connection choices.
- A current Power BI Desktop
- Insights Analyst access to the target Viva Insights partition
- Access to the Viva Insights Analyst Workbench
- Permission to use individual-level data for the approved adoption and enablement purpose
- Power BI workspace permission to publish and configure refresh
The report exposes pseudonymous Person IDs in selected detail paths. Customers remain responsible for approved purpose, access control, retention, row-level security, and sensitivity labeling after loading production data.
Use all available Cowork history beginning with the first week Cowork data exists. Aim for at least 12 covered weeks when available so the report can use its full stage window. Do not request or interpret pre-Cowork weeks as zero.
| Setting | Required value |
|---|---|
| Group by | Week |
| Population attributes | Person ID, Organization, Function type, Layer or level, Supervisor indicator |
| Collaboration metrics | Collaboration hours, active connected hours, email hours, chat hours, meeting hours, unscheduled call hours |
| Beyond-hours metrics | After-hours collaboration, weekend collaboration hours, collaboration span |
| Network metrics | Internal network size, external network size, strong ties, diverse ties, network outside organization |
| Setting | Required value |
|---|---|
| Date range | Same period as the Person Query |
| Group by | Day |
| Entities | Person and Service |
| Filter | Service Name equals Cowork |
| Metrics | Session count, Total Copilot Credits used, Spending policy limit, User limit |
The Person Query supplies the full approved population and work-pattern context. The consumption query supplies Cowork activity. The template defensively rechecks the exact Cowork filter, aggregates consumption to person-week, and keeps covered people without Cowork sessions as zero/non-users rather than dropping them.
The animation is the public DecodingSuperUsage Person Query reference. CoworkSuperUser additionally requires the Cowork consumption query above.
Read the complete instructions: Build the Viva Insights queries.
- Run both analyses and wait for Completed or Success.
- Record the shared Partition ID, Person Query ID, and Consumption Query ID.
- Open the selected PBIT and provide the requested values.
- Confirm the expected population, covered weeks, organization values, sessions, and visible work-pattern views.
- Check history-status labels, privacy suppression, and every page for visual errors before publishing.
- Save as PBIX, apply the organization's required Purview sensitivity label, and publish only to an approved Power BI or Fabric workspace.
For Optimized Export, configure an on-premises data gateway after publishing if the consumption CSV remains in a local folder. See Setup and Troubleshooting.
| Resource | Open or download |
|---|---|
| Direct Query Power BI template | CoworkSuperUser - Direct Query.pbit |
| Optimized Export Power BI template | CoworkSuperUser - Optimized Export.pbit |
| Step-by-step setup | SETUP.md |
| Query build instructions | docs/QUERY_SETUP.md |
| Interpretation guide | INTERPRETATION_GUIDE.md |
| Interpretation storyboard | PPTX |
| Narrated walkthrough | MP4 · Transcript · Subtitles |
| Editable PBIP sources | Direct Query · Optimized Export |
| Troubleshooting | docs/TROUBLESHOOTING.md |
| Security guidance | SECURITY.md |
| Release evidence | validation/release-manifest.json |
- Nine report pages
- 39 bookmark-controlled states
- 117 semantic-model measures
- Complete 117-row Methods and Metric Guide
- Start Here saved as the opening page
- Consistent page-header alignment
- Available-history four- and twelve-week windows
- Aggregate person-derived results suppressed below 10 people
- No tenant identifiers, customer data, local QA paths, or cached model data in the distributable templates
Power BI sensitivity labels apply to PBIX files and are not retained in PBIT exports. The public classification is documented here and in the release manifest; customers must label the refreshed PBIX before sharing it.
- The Person Query is the analytical population spine, not proof of Cowork entitlement.
- Observed active-user share is not an eligibility-based adoption rate.
- Sessions are not tasks.
- Credits are not productivity, quality, complexity, time saved, or business value.
- Work-pattern differences are descriptive and non-causal.
- Champion candidates are outreach starting points, not personnel scores.
- Missing optional evidence means unavailable, not zero.
- Pre-Cowork weeks are outside the analysis, not zero-use weeks.
- Aggregate person-derived results below 10 people are suppressed.
Use the interpretation guide and interpretation storyboard before presenting results.
The PBIT files are data-free and contain no customer data or machine-bound
.pbi cache. Production Person Query and consumption results can contain
personal and business information. Never commit them, attach them to an issue,
or place them in an unapproved location.
Define row-level security roles against the Organization table when audiences
should see different functions or organizations. Assign Entra ID security
groups rather than individual accounts and test each role with
Modeling > View as. Workspace admins and semantic-model owners can see all
data.
Read SECURITY.md before using production data.
Before setup, a Viva Insights administrator or Insights Analyst must create two completed analyses in the same partition.
CoworkSuperUser - Direct Query.pbit
CoworkSuperUser - Optimized Export.pbit
CoworkSuperUser Interpretation Storyboard.pptx
INTERPRETATION_GUIDE.md
README.md
SETUP.md
docs/
images/report-pages/
media/
release/
src/
tools/
validation/
- Cowork Adoption Intelligence
- Decoding Super Usage
- Cowork Billing
- Microsoft Analytics Hub
- Viva Insights Analyst Workbench
- Viva Insights Python library
- Viva Insights R library
See ATTRIBUTION.md for the Microsoft patterns and public assets adapted by this project.
The current public release is v1.0.1. Review the changelog and release manifest before broad distribution.
- Use GitHub Issues for reproducible defects and documentation gaps.
- Do not attach tenant exports, credentials, customer identifiers, or identifiable screenshots.
- Star the repository for discovery and watch releases for updated templates, definitions, and walkthrough assets.
This project is licensed under the MIT License.
This project may contain Microsoft trademarks or logos. Use of Microsoft trademarks or logos must follow Microsoft's Trademark and Brand Guidelines. Modified versions must not cause confusion or imply Microsoft sponsorship.