Business Intelligence Foundations and Data Modeling · Lesson 1

What Is Business Intelligence? From Business Question to Traceable Decision

5 min read

Lesson objectives

  • Frame an analyzable business question and connect it to a user and possible decision.
  • Distinguish the measure, dimension, and context needed to interpret a result.
  • Build a simple trace from question to data, calculation, presentation, and decision.

Start with the question, not the chart

Business intelligence is not simply a screen full of numbers. It is an organized way to connect a business question with suitable data, turn those data into interpretable measures and comparisons, and present them in context so a defined user can make a decision or follow an action. The process starts with the need to understand: What are we trying to know? Who will use the result? What decision might change if the signal changes? In that sequence, a chart is one tool in a reasoning chain rather than the goal itself.

From question to traceable decision

A BI path can be decomposed into six connected steps: frame the question, define the unit and context, identify required data, check data quality, calculate measures, and present the result with enough context to support a decision. “Traceable” here means the user can move backward from a number to its definition, source, and limits: What are the numerator and denominator? Which period? Which requests were included? When were the data refreshed? These questions prevent a number from becoming an unexplained fact whose construction nobody can reconstruct.

LayerQuestion we askMaintenance-case example
Business questionWhat do we need to understand?Is delay concentrated in a particular service type?
DataWhich records and fields are needed?Request, type, team, open and close timestamps
MeasureWhich calculation answers it?Share of requests closed within the target time
ContextWhen, where, and for whom?Week, team, service type
DecisionWhat might we do?Rebalance workload or inspect a bottleneck step

There is a useful distinction between a measure and a dimension. A measure often answers “how much?” or “what proportion?” such as request count or average handling time. A dimension answers “by what?” such as team, service type, or week. This wording is a practical aid rather than an absolute rule. “Average handling time = 42 minutes” is less useful than “average handling time by service type over the last four weeks.” The second statement identifies the comparison and creates a path toward explanation.

Context and limits are part of the number

A useful metric has a definition another person can apply in the same way. For example: “target-time closure rate = eligible requests closed within target time ÷ all eligible requests closed in the period.” The definition must also state what “eligible” means, the period, the target duration, and the timestamp source. If cancelled requests or requests waiting for customer approval are excluded, say so. A changed exclusion rule can change the result even when operational performance does not, which is why business reporting needs definitions, not just numbers.

Lesson summary

BI starts with a question and decision, not a dashboard. A number needs a definition, denominator, period, exclusions, and context, and becomes more useful when we know who will use it and what action might change.