I85PREP
⌘ K
Google Business Intelligence study materials
Google Business Intelligence study resources

Google Business Intelligence practice questions

One question at a time. Build your Google Business Intelligence understanding.

First 50 of 200 questions. Each one lists the choices, the correct answer, and an explanation. The checker at the top of this page saves progress in this browser when it is available.

Continue in the practice tool above to load more questions, 50 at a time.

Question 1Foundations of Business Intelligence

In a business intelligence context, which statement best describes the primary purpose of a data warehouse?

  1. It stores only unstructured documents such as emails and images for compliance archiving.
  2. It captures individual transaction records at the moment they occur and enforces operational business rules.
  3. It replaces the need for any data modeling by letting analysts query raw files directly.
  4. It consolidates data from multiple source systems into a structure optimized for analytical querying and reporting. Correct answer.

Explanation

A data warehouse integrates data from disparate operational sources and organizes it for analysis, typically using dimensional or normalized models tuned for read-heavy reporting. Transaction capture and rule enforcement belong to operational systems, not the warehouse. Warehouses still rely on modeling; they do not eliminate it.

Question 2Foundations of Business Intelligence

A retail company wants to know which store locations exceeded their monthly sales targets. Which type of BI output most directly answers this question?

  1. A machine learning model that predicts next year's inventory shrinkage.
  2. A data catalog entry describing the sales table's schema.
  3. A dashboard comparing actual sales against target by store for the selected month. Correct answer.
  4. A nightly extract of raw point-of-sale logs delivered as flat files.

Explanation

A dashboard presenting actual versus target values by store directly surfaces performance against goals and supports the comparison the business needs. Raw extracts require additional processing before any comparison is possible. Predictive models and catalog metadata serve different purposes than performance monitoring.

Question 3Foundations of Business Intelligence

Which description correctly distinguishes a data lake from a data warehouse?

  1. A data lake only accepts relational tables, while a warehouse accepts JSON and log files.
  2. A data lake stores raw data in native formats at scale, while a warehouse stores modeled, processed data for structured analytics. Correct answer.
  3. A data lake enforces strict schema on write, while a warehouse always uses schema on read.
  4. A data lake is used only for real-time dashboards, while a warehouse is used only for machine learning.

Explanation

Data lakes typically land raw data in its native format and apply structure when it is read, supporting broad data types at scale. Warehouses generally hold cleaned, modeled data optimized for SQL analytics. The other descriptions invert these characteristics or assign overly narrow use cases.

Question 4Foundations of Business Intelligence

What is the main role of the ETL process in a BI pipeline?

  1. It defines access-control policies for dashboard viewers.
  2. It extracts data from sources, transforms it into a consistent analytical form, and loads it into a target store. Correct answer.
  3. It visualizes aggregated metrics for executives through interactive charts.
  4. It schedules email delivery of recurring reports to stakeholders.

Explanation

ETL covers the movement and reshaping of data: extraction from source systems, transformation for consistency and quality, and loading into a warehouse or similar target. Visualization, access control, and report distribution are separate downstream or governance functions.

Question 5Foundations of Business Intelligence

An analyst needs a single trusted number for 'monthly active customers' used across several teams. Which practice best supports this goal?

  1. Store the metric calculation in a personal spreadsheet shared by email.
  2. Let each team calculate the metric independently from raw tables.
  3. Publish the metric only as a screenshot in a slide deck.
  4. Define the metric once in a governed semantic layer and have reports reference that definition. Correct answer.

Explanation

A governed semantic layer centralizes metric definitions so every report derives the same value from the same logic, which reduces inconsistency. Independent calculations and personal spreadsheets create divergent results. A static screenshot cannot be reused or audited as a definition.

Question 6Foundations of Business Intelligence

Which item is an example of structured data commonly used in BI reporting?

  1. A collection of customer service call recordings.
  2. A set of free-text product reviews with no predefined fields.
  3. A relational table of orders with columns for order ID, date, and amount. Correct answer.
  4. A folder of scanned invoices stored as image files.

Explanation

Structured data conforms to a predefined schema, such as rows and columns in a relational table, which makes it straightforward to query and aggregate. Images, audio, and free text lack that predefined tabular structure and are typically treated as semi-structured or unstructured.

Question 7Foundations of Business Intelligence

What does a star schema consist of?

  1. A chain of normalized tables with no fact table.
  2. A set of tables linked only by many-to-many bridge relationships.
  3. A single wide table with no relationships to other tables.
  4. A central fact table joined to surrounding dimension tables. Correct answer.

Explanation

A star schema places a fact table containing measures and foreign keys at the center, with dimension tables providing descriptive context around it. This shape simplifies joins and speeds analytical queries. Fully normalized chains describe snowflake or OLTP designs, and a lone wide table is not a star schema.

Question 8Foundations of Business Intelligence

A BI team wants to track how a customer's loyalty tier changed over several years. Which modeling technique is designed for this need?

  1. A slowly changing dimension that preserves historical attribute values. Correct answer.
  2. A fact table that stores only the current tier value.
  3. A fully normalized dimension with no history columns.
  4. A calculated field that overwrites the prior tier on each load.

Explanation

Slowly changing dimensions retain prior attribute values so reports can reflect the state of a dimension at a point in time. Overwriting values or storing only the current tier loses history. A normalized dimension without history columns cannot answer the question either.

Question 9Foundations of Business Intelligence

What distinguishes ELT from ETL in modern cloud BI architectures?

  1. Loading occurs before extraction and transformation in a single step.
  2. Extraction is skipped entirely because data is already clean.
  3. Transformation happens after loading into the target system rather than before loading. Correct answer.
  4. Transformation is performed manually by dashboard viewers.

Explanation

In ELT, raw data is loaded into a target such as a cloud warehouse and transformed there using the warehouse's compute. ETL instead transforms data before loading. Extraction is still required in both approaches, and viewers do not perform transformation.

Question 10Foundations of Business Intelligence

A dashboard shows a revenue figure that differs from the finance team's official report. Which action best addresses the root cause in a BI governance context?

  1. Add more charts to the dashboard so users can eyeball the difference.
  2. Restrict access so only one team can view revenue figures.
  3. Compare the metric definitions, filters, and source tables used by each report and reconcile them through a governed definition. Correct answer.
  4. Increase the dashboard refresh frequency until the numbers match.

Explanation

Discrepancies usually stem from differing metric logic, filters, or source systems, so reconciling definitions through governance resolves the underlying cause. Refresh frequency affects recency, not the calculation logic. Adding visuals or restricting access does not fix the definitional mismatch.

Question 11Foundations of Business Intelligence

Which scenario most clearly calls for real-time or streaming BI rather than a nightly batch refresh?

  1. Reviewing last quarter's sales by region during an annual planning meeting.
  2. Analyzing year-over-year customer churn trends.
  3. Publishing a monthly headcount report to HR leadership.
  4. Monitoring fraud alerts that must surface within seconds of a suspicious transaction. Correct answer.

Explanation

Fraud detection requires low-latency delivery because action must be taken within seconds, which is a streaming use case. Quarterly, monthly, and year-over-year analyses tolerate batch refreshes since decisions are not time-critical at the second level.

Question 12Foundations of Business Intelligence

What is the purpose of a surrogate key in a dimension table?

  1. To measure the numeric performance of the dimension over time.
  2. To store the human-readable business identifier exactly as it appears in source systems.
  3. To encrypt sensitive customer attributes within the dimension.
  4. To provide a system-generated, stable identifier that is independent of source business keys. Correct answer.

Explanation

Surrogate keys are warehouse-generated integers that uniquely identify dimension rows and remain stable even when source business keys change or are reused. They decouple the warehouse from source key volatility. Encryption and numeric measurement are unrelated to the role of a surrogate key.

Question 13Foundations of Business Intelligence

A BI developer needs to combine a large fact table with a small lookup table that has duplicate rows for some keys. What is the most appropriate first step?

  1. Deduplicate the lookup table to one row per key before joining so row counts are not inflated. Correct answer.
  2. Join directly and accept the multiplied rows as valid output.
  3. Convert both tables to unstructured files before joining.
  4. Drop the fact table and report only from the lookup table.

Explanation

Joining to a lookup with duplicate keys multiplies fact rows and inflates measures, so resolving duplicates to one row per key preserves correct aggregation. Accepting inflated rows produces wrong totals. Dropping data or converting to unstructured files does not solve the join problem.

Question 14Foundations of Business Intelligence

Which statement best explains why data quality checks are placed early in a BI pipeline?

  1. Data quality checks are only relevant to unstructured data.
  2. Catching defects at ingestion prevents flawed data from propagating into models and reports that many users rely on. Correct answer.
  3. Early checks guarantee that business definitions never change.
  4. Early checks eliminate the need for any downstream validation.

Explanation

Detecting issues near the source limits the blast radius, since bad records would otherwise flow into transformations, models, and published reports. Early checks complement rather than replace downstream validation. They apply to all data types and do not freeze business definitions.

Question 15Foundations of Business Intelligence

In a BI context, what does data lineage describe?

  1. The order in which dashboard users are granted access permissions.
  2. The physical storage location of backup tapes in a data center.
  3. The chronological list of meetings held during a project.
  4. The documented path of a data element from its source systems through transformations to its final reports. Correct answer.

Explanation

Lineage traces how data moves and changes across systems, which supports impact analysis, debugging, and trust in reported metrics. Storage locations, permission ordering, and project meetings are unrelated to the flow of data through transformations.

Question 16Foundations of Business Intelligence

A team must choose between a normalized schema and a dimensional schema for a reporting warehouse. Which consideration most favors the dimensional approach?

  1. Analysts frequently run aggregate queries and benefit from simpler joins and faster reads. Correct answer.
  2. The system must support thousands of concurrent order-entry transactions per second.
  3. The schema must avoid any redundancy at all costs.
  4. Storage cost must be minimized regardless of query complexity.

Explanation

Dimensional models denormalize descriptive attributes into dimensions, which reduces join complexity and improves read performance for aggregate reporting. High-concurrency transaction processing and strict redundancy elimination favor normalized designs instead.

Question 17Foundations of Business Intelligence

Which metric type is additive across all dimensions in a fact table?

  1. An average order value.
  2. A percentage of customers who churned.
  3. A sales amount that can be summed across stores, products, and dates. Correct answer.
  4. A ratio of returns to total orders.

Explanation

Fully additive measures such as sales amount can be summed across every dimension without distortion. Ratios, percentages, and averages are non-additive or semi-additive and must be recomputed from their underlying components rather than summed.

Question 18Foundations of Business Intelligence

What is the primary function of a data catalog in a BI environment?

  1. It renders dashboards for end users in a browser.
  2. It inventories data assets with metadata such as descriptions, owners, and usage so users can find and trust them. Correct answer.
  3. It stores raw event streams for real-time processing.
  4. It executes the transformation logic that populates the warehouse.

Explanation

A catalog is a metadata inventory that documents what data exists, who owns it, and how it is described, helping analysts discover and assess assets. Transformation engines, visualization tools, and streaming stores are separate components with different functions.

Question 19Foundations of Business Intelligence

A company stores all raw events in a lake and now wants reliable, governed reporting on curated business metrics. Which approach best fits this need?

  1. Delete the lake and rely on ad hoc spreadsheets instead.
  2. Allow each analyst to define their own metric logic against the raw files.
  3. Point every report directly at the raw event files.
  4. Build a modeled warehouse or curated layer on top of the lake and publish governed metrics from it. Correct answer.

Explanation

A curated layer or warehouse on top of the lake provides modeled, governed data that supports consistent reporting while preserving the lake's raw storage. Reporting directly on raw files or letting analysts define metrics independently reintroduces inconsistency and quality risk.

Question 20Foundations of Business Intelligence

An analyst notices that a dashboard's monthly totals match the source system but yearly totals are lower than expected. Which cause is most consistent with this pattern?

  1. The visualization uses a color palette with poor contrast.
  2. The date dimension has gaps or duplicates that distort aggregation at higher levels. Correct answer.
  3. The users lack permission to view monthly figures.
  4. The dashboard refresh frequency is set too low.

Explanation

When lower-level totals reconcile but higher-level aggregation does not, the problem usually lies in the date dimension's completeness or uniqueness, which affects period rollups. Refresh frequency, color choices, and permissions would not selectively distort yearly totals.

Question 21Foundations of Business Intelligence

A BI architect must support both historical trend reporting and near-real-time operational monitoring from the same data platform. Which design is most appropriate?

  1. Use only nightly batch loads for both requirements.
  2. Use only streaming for all reporting, including multi-year trends.
  3. Use a batch layer for historical modeled reporting and a streaming path for low-latency operational metrics, with consistent definitions across both. Correct answer.
  4. Maintain two entirely separate platforms with independently invented metric definitions.

Explanation

A hybrid design lets batch processing serve deep historical analysis while a streaming path meets low-latency monitoring, and shared definitions keep the two consistent. Batch-only cannot meet real-time needs, streaming-only struggles with large historical scans, and independent definitions create conflicting numbers.

Question 22Foundations of Business Intelligence

A warehouse team adds a new source system whose customer IDs overlap numerically with existing IDs but represent different people. What is the most robust modeling response?

  1. Assign surrogate keys in the customer dimension and map each source's business key to the correct surrogate key. Correct answer.
  2. Concatenate the two ID values into a single string without mapping.
  3. Join the two sources directly on the numeric customer ID.
  4. Drop one source to avoid the conflict.

Explanation

Surrogate keys insulate the warehouse from source key collisions by giving each real-world entity a unique internal identifier, with crosswalk mapping from each source key. Joining on the colliding numeric ID would mix distinct customers, and discarding data or concatenating IDs without mapping does not resolve identity correctly.

Question 23Foundations of Business Intelligence

Which situation most strongly indicates that a metric should be defined in the semantic layer rather than in individual reports?

  1. Multiple teams report the same KPI with slightly different filters and produce conflicting values. Correct answer.
  2. A single analyst needs a one-off exploratory calculation.
  3. A report requires a custom color scheme.
  4. A dashboard needs a new title font.

Explanation

When several teams compute the same KPI differently, centralizing the definition in a semantic layer enforces one calculation and filter logic, eliminating conflicts. One-off exploration, color schemes, and fonts are presentation or ad hoc concerns that do not require governed metric definitions.

Question 24Foundations of Business Intelligence

In a BI pipeline, what is the difference between a fact table and a dimension table?

  1. A fact table and a dimension table always contain identical columns.
  2. A fact table holds only text fields, while a dimension table holds only numeric fields.
  3. A fact table stores descriptive attributes, while a dimension table stores numeric measures.
  4. A fact table stores measurable events and foreign keys, while a dimension table stores descriptive attributes that provide context. Correct answer.

Explanation

Fact tables capture quantitative events and reference dimensions through foreign keys, whereas dimension tables hold the descriptive context used to filter and group those measures. The other options invert or misstate these roles.

Question 25Foundations of Business Intelligence

Which activity is part of data profiling during BI development?

  1. Writing the project's communication plan.
  2. Examining a dataset to assess value distributions, null rates, and key uniqueness. Correct answer.
  3. Designing the dashboard's layout and navigation.
  4. Configuring single sign-on for BI users.

Explanation

Data profiling analyzes a dataset's structure and content, such as null rates, distinct values, and key uniqueness, to inform modeling and quality rules. Dashboard layout, authentication, and project communication are separate concerns.

Question 26Foundations of Business Intelligence

Which description best captures what business intelligence means within an organization?

  1. The practice of writing low-level code to optimize database storage engines
  2. The set of legal controls governing how employee records are retained
  3. The hardware procurement process for provisioning on-premises servers
  4. The process of converting raw data into actionable insights that support business decisions Correct answer.

Explanation

Business intelligence centers on turning raw data into insights that inform decisions, combining processes, people, and technology. It is not primarily about low-level engineering, legal retention policy, or hardware purchasing.

Question 27Foundations of Business Intelligence

A BI team is choosing between a descriptive dashboard and a predictive model. Which pair correctly matches each capability to its purpose?

  1. Descriptive reporting detects anomalies only; predictive analytics stores raw transactional records
  2. Descriptive reporting prescribes the best action; predictive analytics summarizes historical trends
  3. Descriptive reporting summarizes what happened; predictive analytics estimates what is likely to happen Correct answer.
  4. Descriptive reporting estimates future outcomes; predictive analytics explains past performance

Explanation

Descriptive analytics explains past and current performance, while predictive analytics uses models to estimate future outcomes. Prescriptive analytics, not descriptive, recommends actions.

Question 28Foundations of Business Intelligence

Which statement best describes the role of a key performance indicator in a BI program?

  1. It is a raw database column that stores the primary key of a transaction record
  2. It is a visualization style used to render geographic maps in dashboards
  3. It is a quantified measure tied to a strategic objective that teams monitor to gauge success Correct answer.
  4. It is a security policy that restricts access to sensitive personnel data

Explanation

A KPI is a measurable value linked to a strategic goal, giving teams a way to judge whether they are succeeding. It is a business measure, not a database key, chart type, or access-control rule.

Question 29Foundations of Business Intelligence

Which group of users is typically served by a BI dashboard rather than by ad hoc SQL queries?

  1. Database administrators who tune query execution plans
  2. Business stakeholders who need recurring, standardized views of performance metrics Correct answer.
  3. Network engineers who monitor switch throughput
  4. Application developers who write stored procedures for transactions

Explanation

Dashboards deliver recurring, standardized metric views to business users who need consistent answers without writing queries. Technical tuning and infrastructure monitoring are operational tasks served by other tooling.

Question 30Foundations of Business Intelligence

What is the main purpose of a data warehouse in a BI architecture?

  1. To manage employee authentication across corporate applications
  2. To execute the online transaction processing that runs the order entry system
  3. To host the public-facing web application and its session state
  4. To consolidate and store integrated, query-optimized data for analytical reporting Correct answer.

Explanation

A data warehouse consolidates data from multiple sources and structures it for analytical query performance. Transaction processing, web hosting, and identity management are handled by other systems.

Question 31Foundations of Business Intelligence

Which statement accurately describes a data mart?

  1. A tool that schedules extraction jobs across source systems
  2. A staging area that holds only raw unprocessed event logs
  3. A subset of warehouse data scoped to a specific business area or team Correct answer.
  4. A metadata repository that stores only column definitions

Explanation

A data mart is a focused slice of warehouse data serving one business domain, such as finance or marketing. Staging areas, metadata repositories, and schedulers are different components.

Question 32Foundations of Business Intelligence

Which practice best reflects data governance in a BI environment?

  1. Allowing each analyst to define metric formulas independently
  2. Refreshing dashboards only when a user submits a support ticket
  3. Restricting all documentation to a single engineer's local notes
  4. Defining ownership, access rules, and quality standards for shared data assets Correct answer.

Explanation

Governance establishes ownership, access policy, and quality standards so shared data is trusted and consistently used. Fragmented definitions, undocumented local notes, and ticket-driven refreshes undermine that goal.

Question 33Foundations of Business Intelligence

Which activity is a core responsibility of a BI analyst rather than a data engineer?

  1. Maintaining the cluster infrastructure that runs the warehouse
  2. Translating business questions into visualizations and interpreting the resulting metrics Correct answer.
  3. Configuring replication between database servers
  4. Building ingestion pipelines that land raw source extracts

Explanation

BI analysts focus on interpreting business questions and communicating findings through visualizations. Pipeline construction, cluster maintenance, and replication configuration are engineering responsibilities.

Question 34Foundations of Business Intelligence

A retail manager wants a single view showing daily sales, returns, and average order value refreshed each morning. Which BI deliverable fits best?

  1. A spreadsheet maintained manually by the store clerk
  2. A scheduled dashboard combining those three measures with daily refresh Correct answer.
  3. A one-time ad hoc export of raw transaction rows
  4. A static PDF of last year's annual report

Explanation

A scheduled dashboard delivers recurring, consolidated views of multiple measures with automated refresh, matching the daily need. One-off exports, stale static reports, and manual spreadsheets do not provide reliable recurring insight.

Question 35Foundations of Business Intelligence

An analyst must decide whether a requested chart should show a trend over twelve months or a comparison across five regions. Which choice best matches visualization to the question?

  1. Use a line chart for the twelve-month trend and a bar chart for the regional comparison Correct answer.
  2. Use a single gauge for both because it shows a target
  3. Use a pie chart for both because it displays proportions clearly
  4. Use a scatter plot for both because it shows correlation

Explanation

Line charts reveal change over time, while bar charts support comparison of values across categories such as regions. Pie charts, scatter plots, and gauges answer different questions and would obscure these patterns.

Question 36Foundations of Business Intelligence

A product team asks for a metric that reflects the percentage of active users who return within seven days. Which characterization of this metric is most accurate?

  1. A dimension attribute describing the user's signup channel
  2. A raw additive fact that can be summed across all users
  3. A slowly changing dimension tracking subscription tier
  4. A derived ratio metric that requires defining both the active-user and return populations Correct answer.

Explanation

The metric is a ratio built from two defined populations, so it must be derived rather than summed. It is not a raw fact, a descriptive attribute, or a historical dimension.

Question 37Foundations of Business Intelligence

A BI team is documenting a metric called net revenue. Which element is essential to include in the metric definition?

  1. The calculation logic, included and excluded items, and the source tables used Correct answer.
  2. Only the name of the analyst who requested it
  3. Only the dashboard on which the metric appears
  4. Only the color palette assigned in the visualization

Explanation

A complete metric definition specifies calculation logic, inclusion and exclusion rules, and data sources so consumers interpret it consistently. Dashboard placement, requester name, and styling do not determine meaning.

Question 38Foundations of Business Intelligence

A manager wants to know whether a new checkout flow improved conversion. Which analytical approach best supports this question?

  1. Compare conversion rates before and after the change with a controlled baseline or experiment Correct answer.
  2. List every page view recorded by the web server
  3. Report the total number of sessions during the period
  4. Count the number of developers who worked on the release

Explanation

Establishing whether a change caused an improvement requires comparing outcomes against a baseline or controlled experiment. Raw session counts, page views, and staffing numbers do not isolate the effect of the change.

Question 39Foundations of Business Intelligence

Which statement best distinguishes a leading indicator from a lagging indicator in BI reporting?

  1. A leading indicator is always a dimension, while a lagging indicator is always a fact
  2. A leading indicator refreshes hourly, while a lagging indicator refreshes annually
  3. A leading indicator signals future outcomes, while a lagging indicator confirms results already achieved Correct answer.
  4. A leading indicator confirms past results, while a lagging indicator predicts future outcomes

Explanation

Leading indicators anticipate future performance, whereas lagging indicators report outcomes that have already occurred. The distinction concerns timing of insight, not data model type or refresh frequency.

Question 40Foundations of Business Intelligence

A company wants executives to see high-level KPIs while analysts drill into transaction detail. Which design principle best serves both groups?

  1. Give every user full access to raw transaction tables only
  2. Show only the most granular data to all users to avoid duplication
  3. Remove all filters so every user sees identical output
  4. Layer the reporting so summary dashboards link to progressively detailed views Correct answer.

Explanation

Layered reporting lets executives consume summaries while analysts drill into detail, matching each audience's needs. Forcing all users to raw data or unfiltered output ignores differing information requirements.

Question 41Foundations of Business Intelligence

Which scenario best illustrates the value of a self-service BI capability?

  1. Analysts email spreadsheets to each other for every new question
  2. Business users build filtered views from governed datasets without filing engineering requests Correct answer.
  3. Engineers manually rebuild each report whenever a stakeholder asks a question
  4. Executives wait for the quarterly report to see last quarter's numbers

Explanation

Self-service BI empowers business users to answer questions from governed datasets independently, reducing bottlenecks. Manual rebuilds, spreadsheet exchanges, and long waits for static reports reflect the absence of self-service.

Question 42Foundations of Business Intelligence

What is the primary purpose of establishing a single source of truth for a business metric?

  1. To prevent anyone outside finance from viewing the metric
  2. To store the metric in as many systems as possible for redundancy
  3. To let each department define the metric in its own preferred way
  4. To ensure all reports use one agreed definition so numbers reconcile across teams Correct answer.

Explanation

A single source of truth gives every report the same agreed definition, so figures reconcile across teams. Duplicating definitions or allowing departmental variants recreates the inconsistency it is meant to solve.

Question 43Foundations of Business Intelligence

A BI team notices that two dashboards label the same field as revenue but one includes tax and the other excludes it. Which root cause best explains the discrepancy?

  1. The reports were viewed on different screen sizes
  2. The underlying database engine differs between the reports
  3. The metric definition was never standardized across the two reports Correct answer.
  4. The dashboards were built with different color themes

Explanation

Differing inclusion rules indicate that the metric was defined separately rather than standardized. Styling, database engine, and screen size do not change what a revenue calculation includes.

Question 44Foundations of Business Intelligence

Which question is most appropriate to answer through descriptive BI reporting rather than predictive modeling?

  1. Which regions produced the highest sales last quarter Correct answer.
  2. What price will maximize profit over the coming year
  3. Which customers are most likely to churn next month
  4. How many units will be demanded next season

Explanation

Summarizing which regions led last quarter is a descriptive question about the past. Churn likelihood, future demand, and profit optimization require predictive or prescriptive modeling.

Question 45Foundations of Business Intelligence

A BI initiative must prioritize which requests to build first. Which criterion best aligns prioritization with business value?

  1. Rank requests by the alphabetical order of the requesting department
  2. Rank requests by the decision impact and the number of users affected Correct answer.
  3. Rank requests by how quickly a developer can finish the chart
  4. Rank requests by the size of the underlying database table

Explanation

Prioritizing by decision impact and user reach directs effort toward the highest business value. Alphabetical order, table size, and build speed are not measures of business benefit.

Question 46Foundations of Business Intelligence

An organization wants to reduce time-to-insight for recurring operational questions while keeping metric definitions consistent. Which combination of capabilities addresses both goals?

  1. Ad hoc queries written separately by each analyst plus manual spreadsheet consolidation
  2. A semantic layer with governed metric definitions plus scheduled dashboards for recurring questions Correct answer.
  3. Nightly exports emailed to stakeholders plus individual local spreadsheets
  4. A data lake containing only raw files plus a shared folder of screenshots

Explanation

A semantic layer centralizes agreed definitions, and scheduled dashboards deliver recurring answers quickly, together reducing time-to-insight without sacrificing consistency. Separate ad hoc queries and manual spreadsheets reintroduce inconsistency and delay.

Question 47Foundations of Business Intelligence

A BI program has rich dashboards but stakeholders still distrust the numbers. Which foundational investment most directly addresses trust?

  1. Documented metric definitions, data quality controls, and visible lineage for reported fields Correct answer.
  2. A larger number of dashboards covering every possible question
  3. Additional chart types added to every dashboard page
  4. More frequent full refreshes of every dashboard regardless of need

Explanation

Trust grows when consumers can see how metrics are defined, that data is validated, and where values originate. More chart types, indiscriminate refreshes, and dashboard proliferation add volume without addressing credibility.

Question 48Foundations of Business Intelligence

A team is deciding whether a requested analysis belongs in the BI semantic layer or in a one-off notebook. Which factor most strongly favors the semantic layer?

  1. The result depends on a temporary file that will be deleted tomorrow
  2. The question concerns a personal curiosity with no business owner
  3. The analysis is a single exploratory question that will never be repeated
  4. The metric will be reused across multiple reports and must stay consistent over time Correct answer.

Explanation

Metrics reused across reports and expected to stay consistent belong in the governed semantic layer so definitions are shared. One-off, temporary, or ownerless explorations are better suited to ad hoc analysis.

Question 49Foundations of Business Intelligence

A leadership team wants a balanced view of performance rather than a single financial figure. Which BI approach best fits?

  1. A single revenue gauge updated once per year
  2. A raw transaction listing delivered to each executive
  3. A dashboard showing only website click counts
  4. A scorecard tracking several KPIs spanning financial, customer, and operational dimensions Correct answer.

Explanation

A balanced scorecard presents multiple KPIs across financial, customer, and operational perspectives, giving leadership a well-rounded view. A single gauge, raw listings, or click counts alone are too narrow.

Question 50Foundations of Business Intelligence

A BI team needs to decide where raw event data should land before any cleaning or modeling occurs. Which layer of a modern BI architecture is designed for this purpose?

  1. An operational CRM database used by sales agents daily
  2. A presentation layer that serves curated dashboards to business users
  3. A staging or landing zone that stores source data largely as received Correct answer.
  4. A semantic layer that stores metric definitions and joins

Explanation

A landing or staging zone holds source data close to its original form so downstream transformation can be traced and replayed. Presentation and semantic layers come later in the pipeline, and operational systems are not designed as analytical storage.