What is Business Intelligence? | Google Business Intelligence Certificate

Business Intelligence (BI): Purpose and Business Value

BI is framed as automating processes and information channels to convert relevant data into actionable insights for decision makers.

Core business value themes

• Speed to insight: detect issues/opportunities before they become large problems or before competitors act.
• Near real-time monitoring: dashboards and automated pipelines help leaders react quickly.
• Context + fairness: insights must be interpreted correctly to avoid bias or misleading conclusions.

---

Frameworks / Process Playbooks Mentioned

BI “data maturity” model

• Data maturity = the extent to which an organization can use data to generate actionable insights.
• BI focuses on raising maturity through reporting + ongoing monitoring, not one-time analysis.

BI vs. Data Analytics split (operational distinction)

• Data analysts: answer what happened.
• BI professionals: build tools that enable capture → analyze → monitor and drive continuous improvement.

BI three-stage process (explicit playbook)

• Capture (“what happened”): static/backward-looking data retrieval Example: last month’s purchases.

• Analyze (“why did it happen”): deeper investigation to identify relationships, causes, and predictions.

• Monitor (“what’s happening now”): automated pipelines/data models/dashboards for near real-time detection of changes, risks, and opportunities.

BI strategy model (people–processes–tools–governance)

• People: coordinate cross-functional stakeholders; align vision with business strategy.
• Process: define delivery/support model and user enablement (training + feedback loops).
• Tools: choose dashboards/reports based on user needs; define KPI alignment.
• Governance (not the same as data governance): set BI systems + frameworks, rules/policies.
• Documentation: capture stakeholder requirements + project strategy documents.

BI “toolbox”

• Data model: data organization + relationships
• Data pipeline / ETL: extract–transform–load into a data warehouse/unified systems
• Data visualizations + dashboards: interactive monitoring and storytelling
• Iteration: repeat improvements to drive desired outcomes

Metric selection guidance

• Avoid “vanity metrics”; select metrics that are actionable and informative.
• Apply SMART: Specific, Measurable, Action-oriented, Relevant, Time-bound to refine metrics.

---

Concrete Examples / Use Cases Cited

• National Restaurant Group: analyze millions of customers/seating to optimize food supply and reduce waste.
• Hospital example: integrate data sources to review feedback/outcomes and personalize patient experience.
• Manufacturing: global supply chain data → demand forecasting and inventory optimization.
• Google People Operations: dashboards showing time in the interview journey to determine needs (e.g., hiring more recruiters or schedulers).
• Customer service platform change (case-style example):
  • Project sponsor (director of cloud systems) provides vision/objectives.
  • Systems/software developers enable data flow into tables for BI reporting.
• Abandoned cart (e-commerce):
  • Monitor checkout page loading speed to reduce cart abandonment and retain customers.
  • Visualize customer journey and pinpoint drop-off points.

---

Key Metrics, KPIs, and Targets Mentioned

Goal/target example

• Decrease cart abandonment by 15% in 6 months (illustrates how BI monitoring supports KPI achievement).

KPI vs. metric distinction

• KPI = quantifiable value closely linked to business strategy tracking progress toward goals.
• Metric = a single quantifiable data point evaluating performance; metrics support KPIs.
• KPI = strategic, metric = tactical.

Examples of business KPIs/metrics (categories)

• Customer retention
• Year-over-year sales
• Customer loyalty rate
• Productivity levels
• Monthly profits/losses
• Inventory levels (e.g., pharmacy warehouse)

---

Business Execution Recommendations (Actionable Guidance)

Start with stakeholder alignment

• Proactively identify stakeholders (project sponsor, systems analyst, developer, business stakeholders).
• Ask “detective” questions to uncover what they actually need vs what they request.

Define deliverables early

• Explicitly list deliverables before building: dashboards, reports, analyses, documentation.
• Validate expectations using mock-ups (e.g., sketch dashboard charts).

Build context into reporting

• Include chart titles/legends and explain time periods, entity meanings, and appropriate perspective.
• Use a single shared dashboard as a “central location” to reduce context switching.

Mitigate data quality/availability risks

Data availability challenges:

• Integrity: duplicates, missing data, inconsistent structure, rule violations.
• Visibility: stakeholders don’t know what datasets exist (internal/external repositories).
• Update frequency: sources refresh at different cadences, causing misalignment.
• Change management: system/UI/algorithm changes can silently break logic—plan updates and communications.

Additional guidance:

• Use “good enough is sufficient” while acknowledging limitations (avoid chasing perfection).

Avoid vanity metrics

• Don’t report numbers that “impress” but don’t drive meaningful performance. Example: social follower count without tying to purchases/referrals/revenue.

Dashboard design best practices

• Limit the number of metrics to reduce confusion.
• Align metrics to objectives; avoid vague/high-level indicators.
• Put the most important KPI at the top; group related metrics into sections.

---

Presenters / Sources (Named in Subtitles)

• Sally — Business Intelligence Analyst at Google (program instructor for first course)
• Ed — Product Manager at Google
• Terence — Senior Business Intelligence Analyst
• Christine — Associate Lead Recruiter at Google
• Shopify (source mentioned) — cited for estimate of revenue loss due to cart abandonment (18 billion/year)