Business Intelligence (BI)

2.4 Business Intelligence (BI)

Suggested retrieval lesson: 20–30 minutes.

Recall first

1. Trace the BI pipeline from source data to managerial action.
2. Match descriptive, diagnostic, predictive, and prescriptive analytics to their questions.
3. Why is a beautiful dashboard not necessarily good BI?

Commit before reading.

BI as a decision capability

Business intelligence (BI) is the coordinated use of data, technologies, methods, and people to turn organizational data into insights that support decisions and action. It includes data integration, analytical storage, queries, reporting, visualization, and governance—not only charts. IBM describes BI as using data analysis and business information to support decisions. IBM, “What is business intelligence?”

BI pipeline

Operational/external sources → extract and integrate → clean/transform → warehouse or analytical store → semantic model → analyze → visualize/report → decide and act → monitor feedback

• Extract/integrate: bring together ERP, CRM, spreadsheets, sensors, and external data.
• Clean/transform: resolve duplicates, missing values, units, dates, and definitions; retain lineage.
• Store/model: organize facts and dimensions so users share metric definitions.
• Analyze: query, aggregate, compare, segment, and model.
• Present: reports, scorecards, dashboards, alerts, and narratives suited to the user.
• Act/feedback: implement a decision and compare result with target; revise data, model, or policy.

Bad input, ambiguous definitions, or misleading scales can make downstream BI confidently wrong. Governance covers ownership, quality, access, privacy, and interpretation.

Managers and decisions

Decision type is more important than job title:

• Structured/operational: repetitive rules with known inputs, such as replenishing below a threshold; TPS/MIS reports and alerts help.
• Semi-structured/tactical: some rules plus judgment, such as staffing next month; managers use drill-down, scenarios, and forecasts.
• Unstructured/strategic: ambiguous, long-horizon choices, such as entering a market; BI informs but cannot replace judgment, values, and accountability.

Analytics commonly progresses from descriptive (what happened?) to diagnostic (why?) to predictive (what may happen?) to prescriptive (what action is recommended?). A forecast is not a fact, and a recommendation is not an automatic decision.

Worked example: stockout reduction

A manager sees a dashboard showing 8% stockouts. She drills down by store, product, and day (diagnostic), discovers promotions and supplier lead time are correlated, forecasts next week’s demand (predictive), and compares two reorder policies (prescriptive/what-if). She chooses a policy, monitors stockouts and waste, and feeds actual results back into the pipeline. If the dashboard counted canceled orders as sales, the entire decision would be distorted—so metric definition and lineage must be checked first.

Exercise — reveal after committing

Classify these questions: (a) “How many units did we sell yesterday?” (b) “Why did region B’s margin fall?” (c) “Which customers are likely to churn?” (d) “Which retention action should we offer?”

Revealed answer: (a) descriptive, (b) diagnostic, (c) predictive, (d) prescriptive. The categories can overlap in a real project; the distinction is the question and method, not the visualization used.

Exam lens

• BI ≠ database: a database stores/manages data; BI turns governed data into analysis and action.
• BI ≠ dashboard: a dashboard is one presentation output in a larger pipeline.
• Insight ≠ decision: managers interpret context, constraints, ethics, and risk.
• Mention feedback: outcomes test whether the decision and the BI assumptions were useful.

Rapid revision checklist

• Reproduce the pipeline.
• Explain ETL/ELT, warehouse, semantic model, dashboard, and data lineage.
• Distinguish structured, semi-structured, and unstructured decisions.
• Recall the four analytics questions.
• State why data quality and governance precede visualization.

Key takeaways

1. BI is an end-to-end decision-support capability.
2. The pipeline connects sources to action and feedback.
3. Decision structure determines the appropriate report, analysis, and human judgment.
4. Accurate definitions and trustworthy data matter more than visual polish.

Sources

• Rainer & Prince, Management Information Systems (Wiley) — textbook exam framing for BI, managers, and decision support.
• Laudon & Laudon, Management Information Systems: Managing the Digital Firm, 10th ed. — textbook exam framing for DSS, data warehouses, and analytics.
• Boddy & Boonstra, Managing Information Systems: Strategy and Organization — textbook exam framing.
• IBM, “What is business intelligence?” — supplement for BI capabilities and analysis.
• NIST, Big Data Interoperability Framework, Volume 1 — supplement for data terminology and analytical context.