Top 100 Business Intelligence Interview Questions & Answers for Power BI, Tableau, and Looker Studio — A Complete Guide for Aspiring Business Analysts (12 LPA Ready)
- IOTA ACADEMY
- Oct 14
- 16 min read
In today’s data-driven world, Business Intelligence (BI) professionals play a crucial role in transforming raw data into actionable insights that guide strategic decisions. Whether you’re applying for a Business Analyst role or aspiring to work with top organizations offering competitive packages like 12 LPA, mastering tools such as Power BI, Tableau, and Looker Studio is essential. These platforms have become the industry standard for data visualization, reporting, and analytics due to their ability to simplify complex datasets and communicate insights effectively.
This comprehensive guide — “Top 100 Business Intelligence Interview Questions & Answers for Power BI, Tableau, and Looker Studio” — is designed to help you prepare thoroughly for technical interviews. Each question is curated by industry experts, focusing on real-world concepts, recent updates, and the practical skills employers look for in entry-level and mid-level Business Analysts. The answers are written in a simple, easy-to-understand format with concise explanations that make learning and memorization effortless.
Whether you’re a fresher stepping into the analytics world or a professional looking to strengthen your interview readiness, this guide will help you build confidence, enhance your conceptual clarity, and stand out in BI interviews by demonstrating both technical competence and analytical thinking.
Power BI Questions & Answers
What is Power BI?
Answer: Power BI is Microsoft’s business intelligence tool that lets you connect to various data sources, transform/clean data, build data models, and create interactive visuals & dashboards that can be shared.
Explanation: It integrates ETL (Power Query), modeling (Power Pivot / DAX), visual layer, and sharing (Power BI Service).
What are the core components / layers of Power BI?
Answer: Key components include:
Power BI Desktop (for authoring)
Power BI Service (cloud, sharing)
Power BI Mobile (apps)
Power BI Gateway (connect on-premises data to the cloud)
Power BI Report Server (for on-premises hosting)
Power Query / Dataflows, and the data modeling engine (Vertipaq)
Explanation: These components interplay: you build in Desktop, publish to Service, refresh via Gateway, view on mobile, etc.
What is the difference between “Import” mode and “DirectQuery” mode in Power BI?
Answer:
Import mode: data is loaded into the Power BI internal in-memory engine (fast for queries).
DirectQuery mode: Power BI does not import data; each interaction queries the underlying source.
Explanation / trade-offs: Import offers speed but needs refresh; DirectQuery ensures data is up to date but performance depends on the source. (Also there is Dual / Composite mode combining both in newer versions.)
What is a “Live Connection” mode in Power BI? How is it different from DirectQuery?
Answer: Live Connection is a mode where Power BI connects to an existing semantic layer like an Analysis Services model (or an existing Power BI dataset). Unlike DirectQuery which queries raw tables, Live Connection uses a prebuilt model. You can’t add new modeling on top.
Explanation: Useful when enterprise already has semantic models. Newer updates also allow “Shared datasets / composites” where you connect to another Power BI dataset.
What is Power Query?
Answer: Power Query is the ETL (Extract, Transform, Load) engine in Power BI. You use it to connect to sources, clean, reshape, filter, merge, pivot/unpivot data before loading into the model.
Explanation: It uses the “M” language under the hood. Many transformations are UI-driven (no code), but advanced users can write M scripts.
What is DAX (Data Analysis Expressions)?
Answer: DAX is the formula / expression language used in Power BI (and Power Pivot) for defining calculations, measures, calculated columns, tables.
Explanation: It is somewhat similar to Excel formulas but more powerful (with row context, filter context). Key functions: CALCULATE, FILTER, SUMX, RELATED, etc.
What’s the difference between a measure and a calculated column in Power BI?
Answer:
Calculated Column: Evaluated row by row when the data is loaded (or refreshed). The result becomes part of the table.
Measure: Computed on the fly (when visuals render) using the current filter context.
Explanation: Use columns when you want a persistent value per row; use measures for aggregations and dynamic calculations (sums, averages, ratios).
What is a calculated table?
Answer: A table you create via DAX (like Table2 = FILTER(Table1, ...)). It’s computed and materialized during refresh.
Explanation: Useful for intermediate aggregations, supporting tables, or classification tables.
How do relationships between tables work in Power BI? What are cardinalities?
Answer: You connect tables via relationships (e.g. one-to-many, many-to-many, one-to-one) on key fields. Cardinality describes how many matches each side can have.
Explanation: For example, Customer (1) → Orders (*). Many-to-many is also supported now (via composite models).
Also, filtering propagation (direction) matters; you can have single-direction or bidirectional cross-filtering.
What is cross-filtering direction? What is bi-directional filtering?
Answer: Direction determines how filters on one table propagate to another.
Single-direction: Filter flows only one way.
Bi-directional: Filter flows both ways.
Explanation: Bi-directional filters simplify some cases (e.g. many-to-many relationships), but overusing them can degrade performance or cause ambiguous results.
What is “Ambiguity in relationships” / bidirectional filtering issues?
Answer: If you have a model where filters can flow in multiple paths (loops), you may get ambiguous filter propagation. Power BI disallows ambiguous paths or leads to unpredictable behavior unless clarified.
Explanation: Use directions or inactive relationships to resolve.
How can you optimize performance of Power BI reports / dashboards?
Answer / tips (memorize a few):
Limit visuals on a page
Use fewer calculated columns; prefer measures
Reduce cardinality (e.g. numeric keys instead of text)
Use star schema (denormalize)
Limit use of bi-directional filters
Use query folding (push transformations to source)
Use aggregations / summary tables
Optimize DAX (avoid complex nested filters)
Proper indexing on source DBs
Explanation: Performance is multi-layered (query, model, DAX, visuals).
What is query folding?
Answer: It is the process where Power Query transformations are translated (folded) into native queries (e.g. SQL) so they run in the source system, not inside Power BI.
Explanation: This is more efficient (less data moved). Some transformations break folding; you must check in “View Native Query.”
What is a "star schema" vs “snowflake schema” in data modeling? Which is preferred?
Answer:
Star schema: one central fact table with dimension tables around it (denormalized).
Snowflake schema: dimension tables normalized into further sub-tables.
Preferred: Star schema is preferred in BI models because it simplifies relationships, reduces joins, and improves performance.
Explain row-level security (RLS) in Power BI.
Answer: RLS restricts which rows in the data model a user can see (based on roles). You define security roles in Power BI Desktop (via DAX filters on tables), and assign users to those roles on the Service.
Explanation: Useful for multi-tenant models or restricting data by department, geography, etc.
Can you implement dynamic RLS (e.g. user sees only their region)?
Answer: Yes. You can use DAX functions like USERNAME() or USERPRINCIPALNAME() in the filter expression to dynamically filter data based on the logged-in user.
Explanation: This lets you build one model and have user-specific views.
What is the difference between a dashboard and a report in Power BI?
Answer:
Report: One or more pages of visuals built in Power BI Desktop or Service.
Dashboard: A single canvas in Power BI Service that can pin visuals from multiple reports.
Explanation: Dashboards are for high-level, aggregated insights; reports let deeper exploration.
What is a “dataflow” in Power BI?
Answer: Dataflow is Power BI Service’s ETL/transform feature, allowing you to create reusable data transformations in the cloud (using Power Query) and then use these transformed tables across multiple datasets.
Explanation: Helps with ETL centralization and reuse.
What is a Gateway in Power BI? Why is it needed?
Answer: Gateway acts as a bridge between on-premises data sources and the Power BI cloud service. It securely transfers queries and data updates.
Explanation: Without it, Power BI Service cannot access on-prem databases (unless they are open to internet).
What are different types of gateways?
Answer:
Personal mode: runs under user context; for individual use.
Enterprise / standard gateway: shared across users; supports many data sources.
What is “Incremental refresh” and when should you use it?
Answer: Incremental refresh allows only a subset (e.g. recent partitions) of data to be refreshed instead of full dataset refresh, which saves time and resources.
Explanation: Very useful when datasets are large and historical data doesn’t change.
What is a tooltip in Power BI?
Answer: A tooltip is a mini-popup that shows more details when you hover over a visual (data point). You can customize tooltip visuals (e.g. charts within tooltips).
Explanation: Useful to show extra context or drill-down info without cluttering main visual.
What are custom visuals in Power BI?
Answer: Custom visuals are visuals built by the community or developers (beyond the default visuals). You can import them from AppSource or build your own using the SDK (TypeScript, etc.).
Explanation: Helps to get specialized charts not present by default.
What is “bookmarking” in Power BI?
Answer: Bookmarks are snapshots of the current state of visuals (filters, slicers, drilldowns). You can use them for navigation, storytelling, or to create “toggle” buttons.
What is “drill-through” vs “drill-down”?
Answer:
Drill-down: Within the same visual, you move from higher-level hierarchy (e.g., Year → Month → Day).
Drill-through: Click a data point in one report page and jump to another page filtered to that context.
What is “decomposition tree” visual in Power BI?
Answer: Decomposition Tree allows hierarchical breakdowns of a measure by various dimensions (you choose which dimension to break next). Good for root-cause analysis.
How do you handle missing / null values in Power BI?
Answer: Use Power Query functions (Replace Values, Fill Down/Up, conditional logic) or DAX functions (IF, BLANK, COALESCE). Also filter them out if they don’t make sense.
What is “Q&A” in Power BI?
Answer: Q&A allows users to ask natural language questions (in English) and get instant visuals (e.g. “Total sales by region”). It uses the semantic model.
What is the difference between Power BI Desktop and Power BI Service?
Answer:
Desktop: authoring environment where you build data models, visuals, transformations.
Service: cloud platform for publishing, sharing, collaboration, refreshing, dashboards.
What file formats does Power BI support? (for import / export / templates)
Answer:
.pbix (report + data model)
.pbit (template without data)
You can import from Excel, CSV, JSON, XML, etc.
Export visuals or data to Excel / CSV.
What is data profiling in Power Query?
Answer: Data profiling gives you info about column quality, distribution, distinct count, nulls, etc., helping you understand the data before transformation.
What is bi-directional cross-filtering? What are its risks?
Answer: As in Q10, it allows filters to flow both ways between tables. Risk: can cause ambiguous relationships or performance issues. Use carefully and explicitly.
What are “aggregations” in Power BI?
Answer: Precomputed summary tables (at a higher level of granularity) which reduce the need to scan detail rows. Power BI can switch between aggregation and detail tables automatically.
What is the difference between “ALL” vs “ALLEXCEPT” DAX functions?
Answer:
ALL(table) clears all filters on that table.
ALLEXCEPT(table, column1, column2, ...) clears filters on all columns except the ones you specify (keeps those filters).
What is the use of CALCULATE in DAX?
Answer: CALCULATE modifies filter context to compute a measure with new filters (or remove filters). Many advanced DAX patterns use CALCULATE.
What is the difference between FILTER and a row context vs. CALCULATETABLE?
Answer:
FILTER returns a table after applying the filter expression.
CALCULATETABLE applies filter modifications and returns a table; it's like CALCULATE but returns table instead of scalar.
How would you compute a running total (cumulative sum) in DAX?
Answer: For example:
This sums all amounts up to current date.
What is “time intelligence” in DAX?
Answer: Time intelligence are DAX functions to work with time-based calculations (year to date, same period last year, month to date, etc.). Examples: TOTALYTD, DATESYTD, SAMEPERIODLASTYEAR.
How do you handle slowly changing dimensions (SCD) in Power BI?
Answer: Common methods:
Overwrite (Type 1)
Add versioning columns and historical rows (Type 2)
Use separate dimension tables and join with effective date ranges
Explanation: Use logic in ETL / Power Query or DAX to maintain historical tracking.
What are some limitations of Power BI you should be aware of?
Answer: Some known limitations (as of recent):
Dataset size limits (max size in Service, e.g. 1 GB for Pro)
Complex DAX / heavy bi-directional relationships may degrade performance
Some sources don’t support DirectQuery
Limited export of visuals
For very large enterprises, SSAS / Azure Analysis Services might be preferable
You can add a few more advanced ones (composites, incremental refresh, performance analyzer) based on the company’s tech stack.
Tableau Questions & Answers (approx. 25)
What is Tableau?
Answer: Tableau is a data visualization and analytics tool that allows users to connect to data, create visualizations, dashboards, and share them. It emphasizes ease-of-use, drag-and-drop interface, visual storytelling.
What are different versions / products in Tableau?
Answer: Tableau Desktop (authoring), Tableau Server / Tableau Online (hosting/sharing), Tableau Public, Tableau Prep (for ETL / cleaning), Tableau Mobile.
What is a “data source” in Tableau? What is a “live” vs “extract” connection?
Answer: Data source defines where Tableau gets data (Excel, SQL, etc.).
Live: connects directly and queries in real time.
Extract: pulls a snapshot of data into Tableau’s internal engine (faster performance).
What are dimensions and measures in Tableau?
Answer:
Dimension: categorical field (e.g. Region, Product).
Measure: numeric field (e.g. Sales, Profit).
What is Tableau’s “Show Me” feature?
Answer: A panel that suggests a set of chart types based on selected fields; helps users quickly pick visuals.
What is dual-axis chart / combo chart in Tableau?
Answer: A chart with two axes (e.g. left & right) to plot two measures with different scales, overlaid in same view.
What is a calculated field in Tableau?
Answer: Similar to Excel formula or DAX: you define a new field derived from existing ones (e.g. Profit Ratio = Profit / Sales).
What is level of detail (LOD) expressions in Tableau?
Answer: LOD expressions allow aggregation at different granularities than the view. Types: FIXED, INCLUDE, EXCLUDE.
FIXED [Region]: SUM(Sales) calculates sum at Region level regardless of view.
What is data blending in Tableau?
Answer: A method to join data from two different sources on a common field (like left join) at run time. Primary and secondary data sources concept.
What is a context filter?
Answer: A filter that first reduces the data, making other filters operate on that subset. Useful when you have dependent filters.
What is aggregation vs disaggregation?
Answer:
Aggregation: combining data (sum, average)
Disaggregation: showing row-level data (individual rows)
What is the difference between “discrete” and “continuous” fields in Tableau?
Answer:
Discrete: treated as categories; shown as headers.
Continuous: treated as continuous scale (axis).
What is a dashboard vs story in Tableau?
Answer:
Dashboard: collection of multiple views / sheets in one page.
Story: sequence of dashboards / views telling a narrative with captions.
How do you optimize Tableau dashboard performance?
Answer / tips:
Use extracts instead of live when possible
Limit number of marks, reduce filters
Use context filters / indexing
Limit complex calculations
Use data source optimizations (indexes on source)
How do you share Tableau reports?
Answer: Publish to Tableau Server or Tableau Online; you can embed, share links, control permissions.
What is parameter in Tableau?
Answer: A dynamic value (single value) that users can input or change, used in calculations or filters to make dashboards interactive.
What is “Tableau Prep”?
Answer: A Tableau product for data preparation — cleaning, shaping, joining, pivoting data before analysis.
How do you handle nulls / missing values in Tableau?
Answer: Use calculated fields (IFNULL, ZN), filters to exclude, or “Data Interpreter” (for Excel / CSV) to detect nulls.
Explain a use case of “Parameter + Filter” combination.
Answer: Suppose you let user pick a threshold (via parameter) and then filter data where Sales > [Threshold]. Parameter drives logic.
What is a “Table calculation” in Tableau?
Answer: A calculation applied after query results are fetched; based on table structure, e.g. running total, percent of total, index.
What is partitioning in table calculation?
Answer: You define how to break the data (panel / partition) and direction (address) over which the calculation flows.
What is the difference between joining and blending?
Answer:
Joining: data combined at source / in Tableau between tables in same data source.
Blending: combining aggregated data from two separate data sources (at visualization time).
What is “Extract Filters” vs “Data Source Filters”?
Answer:
Extract filters: applied when generating extract; reduce data in extract.
Data Source filters: applied in both live and extract; filter data in the source pane.
What are performance recording / logging in Tableau?
Answer: You can enable performance recording (Help → Settings) to capture times of queries, rendering, etc., or check log files to identify slow queries.
Recent change / update in Tableau (if known)?
Answer: Tableau’s newer versions have improved features like Set Control (user-driven sets), Explain Data / Ask Data (natural language), Dynamic Parameters (parameters that refresh), and Metrics in server/online. You should check the version used in the company.
Looker / Looker Studio (and LookML) Questions & Answers (approx. 20)
(Here “Looker Studio” often refers to Google Looker Studio (formerly Data Studio), but many interview questions are about Looker + LookML. So I mix them but note distinctions.)
What is Looker (or Google Looker)?
Answer: Looker is a BI / analytics platform (now part of Google Cloud) that enables data exploration, modeling (via LookML), dashboards, and embedding. It directly queries databases (no extraction).
What is LookML?
Answer: LookML is the modeling language in Looker. You define views, dimensions, measures, and relationships. It abstracts SQL so users don’t write SQL.
What is a “view” in LookML?
Answer: A view represents a table (physical or derived). You define dimensions and measures in it.
What is a “model” in LookML?
Answer: A model configuration defines which explores are available to users and points to views; it defines which views can be joined.
What is an “Explore” in Looker?
Answer: Explore is the UI interface where users can drag dimensions/measures to query data; behind the scenes, Looker composes SQL using the model/view definitions.
What are dimensions and measures in LookML?
Answer:
Dimension: a field that you slice or group by (e.g. date, category)
Measure: aggregation over data (sum, count, average)
What are derived tables and PDT (persistent derived tables)?
Answer: In Looker, a derived table is a virtual table created from a SQL query in your LookML project. They are defined within a view file and can perform complex calculations and data transformations that result in a new, logical table. The key difference between a derived table and a Persistent Derived Table (PDT) is how the results are stored and refreshed.
Explain user attributes in Looker.
Answer: User attributes are metadata tied to user accounts (like region, department) that can dynamically drive filtering or access (e.g. sql_where: ${region} = {% user_attribute region %})
What is development mode in Looker?
Answer: It allows developers to make changes in LookML files, test them without affecting production, and then deploy to production.
How do you schedule Looks / dashboards?
Answer: In Looker, you can schedule a Look or dashboard to be emailed in a format (PDF, CSV) at a frequency; you configure recipients and delivery settings.
What are Looker Blocks?
Answer: Prebuilt LookML templates or modules (for common business use cases) that can accelerate development (e.g. retail, ad attribution).
What is the difference between self-hosted Looker and Hosted (Cloud) version?
Answer: The main difference is that self-hosted Looker is an instance you manage on your own infrastructure, giving you complete control over hardware and data, while Hosted (Cloud) Looker is a version managed by Google, which offers greater convenience but less control. Self-hosting requires you to handle all infrastructure and maintenance, while the cloud version handles these aspects, providing more agility and reducing overhead.
What SQL dialects does Looker support?
Answer: Looker supports many SQL dialects (50+), depending on the underlying database. The LookML engine adapts generated SQL accordingly.
What is the difference between Looker and Google Looker Studio (formerly Data Studio)?
Answer:
Looker: full BI platform with modeling (LookML), embedding, strong metadata, user governance, optimized for real-time querying.
Looker Studio: Google’s free visualization tool (formerly Data Studio) focused on dashboarding / report visuals from data sources (BigQuery, Google Sheets, etc.).
In interviews, clarify which “Looker” they mean.
How do you handle performance optimization in Looker?
Answer / tips:
Use derived tables / PDTs
Limit joins
Use persistent results / caching
Filter early (in SQL)
Ensure database is indexed properly
Avoid overly complex nested LookML logic
What is a “datagroup” in Looker?
Answer: Datagroups define refresh schedules / triggers for PDTs (e.g. time-based or when source data changes) so that derived tables are kept up to date.
What is access / permission model in Looker?
Answer: Looker uses roles and permissions (view, explore, edit, manage), as well as model-level and field-level access, sometimes in conjunction with user attributes.
What is Looker API?
Answer: Looker exposes a RESTful API to automate tasks: run queries, manage users, schedule runs, manage LookML projects.
Can you embed Looker dashboards into external apps?
Answer: Yes, Looker supports embedding dashboards, visuals, and explores into web applications (with secure access).
Recent change / update in Looker (2025)?
Answer: The Looker platform often adds features around embedded analytics, improvements in caching, analytics actions, and tighter integration with Google Cloud. Check release notes of the particular version.
General / Dashboarding / Analytics / BA Questions (approx. 15)
What is business intelligence (BI)?
Answer: BI refers to tools, processes, and practices used to collect, integrate, analyze, and present business data to support decision-making.
What are key qualities of a good dashboard?
Answer:
Clear & intuitive layout
Focus on key KPIs
Minimal clutter
Use of appropriate visual types
Ability to drill / filter
Responsiveness / performance
Storytelling & context
Which chart types would you use for these use cases: trend over time, distribution, part-of-whole, comparison?
Answer:
Trend: line chart, area chart
Distribution: histogram, box plot
Part-of-whole: pie chart, donut, stacked bar, tree map
Comparison: bar chart, column chart
How would you redesign a cluttered dashboard to make it stakeholder-friendly?
Answer: Steps: group related visuals, reduce number of charts per page, use white space, highlight key metrics, use consistent color scheme, remove redundancy, add filters/slicers, add explanatory titles.
How would you explain a complex metric to a non-technical stakeholder?
Answer: Use plain language, analogies, break down the formula into simpler parts, show the input data and result, use visuals, and iterate.
What is data validation / data quality? How do you ensure data is reliable?
Answer: Data validation is checking for accuracy, consistency, completeness. Ensuring reliability via checks: null / outlier detection, reconciliation with source, sample audits, data profiling.
What is ETL / ELT?
Answer:
ETL: Extract data, Transform it (clean, shape), then Load into target.
ELT: Extract, then Load raw data into target, then Transform inside the target (e.g. DB).
What is correlation vs causation? Why is it important in analytics?
Answer:
Correlation: two variables move together (statistical relation)
Causation: one causes the other
It’s important not to imply causation just because variables correlate; you must investigate deeper (experiments, domain knowledge).
How would you handle conflicting metrics (e.g. two departments define “customers” differently)?
Answer: Facilitate alignment: bring stakeholders together, define a canonical metric definition, document assumptions, agree on version for dashboard, adjust source if needed.
What is A/B testing? How would you visualize results?
Answer: A/B testing is comparing two variants (A vs B) to see which performs better. Visualize using bar charts with confidence intervals, difference over time, cumulative lift curves.
What is “self-service BI”? What are its pros & cons?
Answer: Self-service BI means business users can generate reports / dashboards without needing IT or BI developers.
Pros: agility, faster insights, empowerment
Cons: risk of inconsistent metrics, uncontrolled data sprawl, performance issues
How do you prioritize which dashboards or reports to build first?
Answer: Use business impact, stakeholder value, effort & complexity, data readiness. Start with high ROI, release iteratively.
What is “data storytelling”?
Answer: The practice of structuring and presenting data (via visuals, narrative, context) to guide stakeholders to an insight or decision.
Name a recent BI / analytics trend.
Answer:
Augmented analytics (AI / ML-based insight suggestions)
Natural language querying
Embedded analytics (dashboards inside apps)
Real-time / streaming dashboards
If you were given a dataset (say, sales over 5 years), what steps would you take from raw to dashboard?
Answer (process):
1. Understand stakeholder requirements / metrics
2. Explore raw data, profiling, understand schema
3. Clean, transform, deal with missing / duplicates
4. Build data model (dimensions, facts, relationships)
5. Define calculations / measures
6. Build visuals / dashboard prototype
7. Get stakeholder feedback & iterate
8. Optimize performance, deploy, schedule refresh
9. Monitor usage and adjust
Tips for using these questions:
Don’t just memorize the answers; try to rephrase with your own words and give real or mock examples
Be ready to dive deeper—if they ask “why?” after your answer, expand
For any tool (Power BI, Tableau, Looker), know which version your prospective employer uses; some features may differ
Use sample data sets during preparation to practice building dashboards
Use visual diagrams if asked (draw relationships, filters, flows)
Looking to enhance your database skills? Join Iota’s Courses and master Data Handling and visualization and more! 🚀
Comments