SecondBrain/20 Work/Ideas/# Organizational Data & Analytics Operating Model.md

Date: 2026-07-22
Status: 💡 Idea

The Idea

Build a company-wide Data & Analytics capability where trusted data is directly accessible to authorized departments, while each business unit retains ownership and expertise over its own business domain.

The Data & Analytics team would not "own all company data." Instead, it would provide the common capability that makes organizational data:

• Accessible

• Trusted

• Consistent

• Integrated

• Documented

• Governed

• Understandable

• Actionable

Each department would retain its business expertise and ownership, while working closely with a dedicated member of the Data & Analytics team who understands that department, its website, projects, objectives, and data.

The long-term goal is not simply to build more Power BI reports.

It is to build the organization's measurement system, allowing departments and leadership to understand organizational performance from trusted data.

Problem / Opportunity

The current organization has a relatively new data culture and historically operates through departmental silos.

There is currently ambiguity around:

• Who owns data

• Who can access data

• Who interprets data

• Who defines KPIs

• Who distributes information

• Which reports are authoritative

• Who decides whether performance is good or bad

Marketing has historically positioned itself as an intermediary for organizational analytics. Departments may be expected to go through Marketing to obtain or interpret their own performance data.

This creates a risk of information gatekeeping.

For example, even when centralized reports are available to departments directly, Marketing may prefer that departments request information through Marketing first.

This model creates unnecessary dependencies:

Department
    ↓
Marketing
    ↓
Data / Reports
    ↓
Marketing interpretation
    ↓
Department

It also creates a potential conflict when data reveals uncomfortable results.

A recent cross-platform video analytics report made previously difficult-to-see performance information directly visible. Some departments were resistant to having poor performance exposed to upper management.

This demonstrates why business ownership should not automatically mean control over whether organizational performance information can be seen.

Another issue became clear when the VP asked:

"What are the KPIs?"

The report successfully solved the data availability problem, but the organization had not yet clearly established which metrics actually indicate whether video performance is succeeding or failing.

This represents an opportunity to move beyond reporting toward true Data Governance and performance measurement.

Why It Matters

A strong Data & Analytics operating model would reduce departmental silos and make trusted information directly available to the people who need it.

Instead of asking:

"Who controls the numbers?"

the organization should be able to ask:

"What do the numbers tell us, and what should we do about them?"

The business value includes:

• Better executive decision-making

• Greater transparency

• Less information gatekeeping

• Consistent KPI definitions

• Increased trust in reports

• Faster access to information

• More effective self-service analytics

• Better collaboration between Data and business departments

• Better understanding of organizational performance

• Reduced duplication of reports and analysis

• Less dependence on individual departments to distribute information

• Better accountability when performance is poor

• Cross-departmental visibility

The long-term objective is for leadership to understand the health of the entire organization using trusted data.

Who Benefits?

• Executive leadership

• News

• Education

• Marketing

• Finance

• HR

• IT

• Legal

• Digital / Product

• Other business departments

• Data & Analytics team

Ultimately, every department should benefit from having direct access to trusted information and an analytical partner who understands its business.

Possible Approach

1. Business Domains Own Their Business

Departments retain ownership of their respective business domains.

Examples:

HR          → HR / Employee domain
Finance     → Financial domain
Marketing   → Marketing domain
News        → News domain
Education   → Education domain
IT          → Technology / Operations domain
Legal       → Legal / Compliance domain

Business ownership means departments understand their operations and participate in defining what their business metrics mean.

It does not automatically mean that they control who may see organizational performance information.

Ownership, access, reporting, interpretation, and governance are separate responsibilities.

2. Data & Analytics Provides the Shared Capability

The Data & Analytics function should be responsible for establishing and maintaining the organization's shared analytics capability.

Responsibilities could include:

• Data integration

• Analytics architecture

• Power BI / reporting

• Data quality

• Data lineage

• Metadata

• KPI documentation

• Shared definitions

• Cross-domain analytics

• Analytical methodology

• Self-service analytics

• Data access implementation

• Data governance standards

The objective is not:

"All data belongs to Data & Analytics."

Instead:

Business units own their business domains. Data & Analytics makes organizational data trustworthy, integrated, consistently measured, understandable, and accessible to authorized users.

3. Dedicated Data / Technology Partners

Each department already has a dedicated team member who works closely with it on its website, projects, technology, and evolution.

This relationship should evolve into an embedded Data / Technology partnership.

                   Data & Analytics
                         │
          ┌──────────────┼──────────────┐
          │              │              │
      Team Member    Team Member    Team Member
          ↕              ↕              ↕
        News         Education      Other Dept.

These team members should develop enough domain knowledge to understand and interpret the department's analytics.

They should not merely provide numbers.

They should be able to:

• Explain what happened

• Identify patterns

• Investigate changes

• Challenge assumptions

• Understand the department's digital environment

• Connect technical events with analytics

• Recommend further investigation

• Work with departmental experts to understand why something happened

The business department still retains accountability for business decisions.

4. Shared Interpretation

Interpretation should not automatically belong to Marketing.

Instead:

Data & Analytics asks:

What happened?

Where did it happen?

What patterns exist?

Is the data reliable?

Data + Department together ask:

Why did it happen?

The department contributes business context because it understands its operations.

The embedded Data/Technology partner contributes analytical and technical expertise because they understand both the data and the department.

Department / Leadership decides:

Is this performance acceptable?

What should we do about it?

Marketing can contribute Marketing expertise where relevant, but should not automatically become the interpretation layer for News, Education, or other departments.

5. Self-Service by Default

Authorized users should not need to request basic information from Marketing or Data if trusted reports already exist.

The desired model is:

             Trusted Data & Analytics
                      │
        ┌─────────────┼─────────────┐
        ↓             ↓             ↓
      News        Education      Marketing
        ↕             ↕             ↕
  Data Partner   Data Partner   Data Partner

Not:

Department → Marketing → Data → Marketing → Department

The goal is not to replace Marketing as the gatekeeper with Data as the new gatekeeper.

The goal is to remove unnecessary gatekeeping entirely.

6. Establish KPI Governance

Reports should distinguish between:

Metrics

Numbers that describe what happened.

and:

KPIs

Metrics specifically selected to determine whether the organization is succeeding against an objective.

For example, a video analytics report may contain dozens of metrics.

Leadership may ultimately decide that only a small number represent organizational success.

The process should become:

Business Objective
        ↓
Agreed KPI
        ↓
Business Definition
        ↓
Documented Calculation
        ↓
Trusted Data Source
        ↓
Power BI / Analytics
        ↓
Department Interpretation
        ↓
Leadership Decision

Data & Analytics should facilitate this process but should not independently invent corporate KPIs.

Questions to establish for important KPIs:

• What business objective does this KPI measure?

• Who owns its business definition?

• How exactly is it calculated?

• Which systems contribute data?

• What are its limitations?

• How frequently is it refreshed?

• Who may access it?

• Who approves changes?

• Where is the authoritative version?

• What supporting metrics explain changes?

7. Position Marketing's Data Analyst Correctly

Marketing having a Data Analyst is not inherently a problem.

The important question is the analyst's mandate.

A Marketing Data Analyst could appropriately specialize in:

• Marketing campaigns

• Acquisition

• Audience segmentation

• Marketing effectiveness

• Advertising performance

• Marketing KPIs

The concern would be if the role evolves into:

The organizational analyst responsible for interpreting News, Education, Digital, Streaming, and other departments.

Marketing should be one business domain within the broader Data & Analytics ecosystem, not the mandatory gateway to organizational information.

Do not make the organizational argument:

"Marketing shouldn't have an analyst."

Instead establish the broader model in which that analyst naturally becomes a Marketing domain specialist.

8. Expand Beyond Audience Analytics

The current video analytics work can become the foundation rather than the final destination.

Long term, gradually integrate authorized information from additional domains.

                 ORGANIZATIONAL HEALTH

 Financial      Audience       People       Operations
     │              │             │              │
 Revenue          Reach        Headcount        SLA
 Budget          Viewing       Turnover        Uptime
 Costs          Retention       Hiring         Delivery
     │              │             │              │
     └──────────────┴──────┬──────┴──────────────┘
                           ↓
                  Executive Decisions

Potential domains include:

Finance

• Revenue

• Budget

• Actual vs forecast

• Costs

HR

• Headcount

• Hiring

• Turnover

• Workforce trends

IT

• SLA

• Incidents

• Uptime

• Project delivery

• Service requests

Legal

• Compliance indicators

• Contracts

• Rights / licensing where applicable

Audience / Content

• Reach

• Consumption

• Engagement

• Retention

• Content performance

The goal is not one enormous Power BI dashboard.

The goal is an interconnected organizational data ecosystem.

9. Respect Sensitive Data Ownership

Expanding into HR, Finance and Legal should not mean requesting unrestricted access to everything.

These departments may legitimately need strict controls around sensitive information.

Instead:

The department retains ownership and appropriate access control while participating in the organization's governed analytics ecosystem.

For example, HR may expose approved workforce metrics without exposing individual salaries or unnecessary personal information.

This reduces resistance and avoids presenting Data & Analytics as an attempt to take control away from departments.

Effort

Estimated complexity: High

This is not primarily a Power BI or technical project.

It requires:

• Organizational change

• Executive sponsorship

• Data Governance

• Departmental cooperation

• Clear responsibilities

• Trust

• Political navigation

• Technical architecture

• Gradual cultural change

Implementation should therefore be incremental rather than presented as one large transformation project.

Risks / Questions

• Marketing currently has strong relationships with the VP and CEO.

• Marketing has already received approval to hire a Data Analyst focused on interpretation without consultation with the existing Data & Analytics function.

• Marketing may perceive direct self-service analytics as reducing its organizational influence.

• Departments may resist transparency when data reveals poor performance.

• HR and Finance may resist participation because of concerns around control and sensitive information.

• Departments may interpret "central Data & Analytics" as an attempt to take ownership of their data.

• KPI ownership is currently unclear.

• Leadership may not yet recognize the distinction between metrics and KPIs.

• There may be no formal Data Governance mandate.

• Data access and business ownership may currently be confused.

• Embedded team members need sufficient analytical and business knowledge to become credible partners.

• Who ultimately resolves disagreements over KPI definitions?

• Who should sponsor Data Governance at the executive level?

• How should cross-department KPIs be owned?

• What should the formal mandate of Data & Analytics become?

Next Step

Use the existing cross-platform video analytics report as the first practical example of this operating model.

Do not initially attempt to restructure the organization or challenge Marketing's analyst position.

Instead, follow up on the VP's question:

"What are the KPIs?"

Work with leadership and the relevant departments to identify:

1. What are our business objectives for video?

2. Which metrics indicate whether we are succeeding?

3. Who owns the business definition of each KPI?

4. How should each KPI be calculated?

5. Which supporting metrics help explain performance?

Document the agreed definitions and implement them consistently in the existing report.

This creates a small, concrete example of Data Governance and demonstrates the value of the broader operating model.

From there, gradually extend the model to additional domains rather than trying to win the abstract argument over "Who is Data?"

The long-term strategy is:

Don't fight to be called Data. Build the organizational Data & Analytics capability until its role becomes self-evident.

Related

• Data Governance

• Data Quality

• Data Lineage

• Semantic Layer

• Business Glossary

• KPI

• Data Owner

• Data Steward

• Power BI

• Data Culture

• Self-Service Analytics

• CTO