view: orders {
dimension: order_id {
type: number
sql: ${TABLE}.order_id ;;
}

measure: total_sales {
type: sum
sql: ${TABLE}.sale_amount ;;
}
}

3. What are Dimensions in looker?

In Looker, Dimensions and Measures are the two core building blocks that define how your data is explored and analyzed. They work together to give you meaningful insights from your datasets.

1. Dimensions

Definition: Dimensions are descriptive attributes or fields in your data that represent categories, identifiers, or raw values.

Purpose: They describe “what” you are analyzing.

Examples:

Customer Name

Product Category

Order Date

Region

SQL Equivalent: Usually map to a column in a database table.

Example LookML Definition:

lookml

• Purpose: They answer “how much” or “how many” about a dimension.

• Examples:

  • Total Sales (SUM(sales_amount))

  • Average Order Value (AVG(order_value))

  • Number of Orders (COUNT(order_id))

• SQL Equivalent: Typically a SQL aggregate function.

5. Explore vs Dashboard in Looker

In Looker, Explore and Dashboard are two different but connected features for working with data. They serve different purposes in the analytics workflow.

1. Explore

Definition:
Explore is an interactive data exploration tool in Looker that allows users to query data in real-time, slice and dice it using dimensions and measures, and create visualizations.

Key Points:

• Used for ad-hoc analysis.

• You choose a starting dataset (called an Explore) and then add filters, dimensions, and measures to build queries.

• You can create temporary visualizations or save them for reuse.

• Flexibility: You can keep modifying filters and fields until you find the exact insight you need.

• Can be saved as a Look or added to a Dashboard.

Example Use Case:
A data analyst wants to see Total Revenue by Product Category in Q1 2025. They open the Explore, select “Product Category” as the dimension and “Total Revenue” as the measure, and filter for the relevant date range.

2. Dashboard

Definition:
A Dashboard is a collection of saved visualizations (Looks) arranged in a single view for monitoring and sharing insights.

Key Points:

• Ideal for reporting and tracking KPIs over time.

• Consists of multiple tiles (charts, tables, text) that pull data from saved Looks or Explores.

• Supports scheduled delivery via email or Slack.

• More static compared to Explore — dashboards are for monitoring, not deep, ad-hoc analysis.

• Can include interactive filters to adjust all tiles at once.

Example Use Case:
A sales manager wants a Sales Performance Dashboard showing metrics like Total Sales, Sales by Region, Top Products, and Monthly Trends — all updated automatically.

Also Read: Looker Tutorial

6. How Looker Connects to Data

Looker is not a database itself — it’s a Business Intelligence (BI) platform that connects directly to your existing databases in real time. Instead of importing data into Looker, it queries the data where it lives using Live Connection.

1. Live Connection Model

• Looker connects to databases through SQL.

• It uses database drivers and credentials to establish a secure connection.

• No data is stored in Looker permanently — queries are run on the database each time you request data.

• This ensures you’re always working with the most up-to-date information.

7. What are the steps to Connect Looker to a Database ?

1. Choose Your Database Type

   • Looker supports a variety of databases like:

     • Cloud Data Warehouses: BigQuery, Snowflake, Redshift

     • Relational Databases: MySQL, PostgreSQL, SQL Server

     • Analytics Databases: Vertica, Teradata, etc.

Set Up Database Credentials

1. • In Looker’s Admin panel, provide:

   • Host (server address)

   • Port (database connection port)

   • Username & Password

   • Database name

2. Select the Dialect

   • Choose the correct SQL dialect so Looker generates queries compatible with your database (e.g., Snowflake SQL, BigQuery Standard SQL, etc.).

3. Test the Connection

   • Looker checks if it can successfully connect and run sample queries.

4. Create Models in LookML

   • Once connected, Looker developers define Explores, Views, Dimensions, and Measures in LookML to structure how the data can be queried.

8.  How Queries Work in Looker ?

• When you run a report in Looker, here’s what happens:

Looker translates your selections (dimensions, measures, filters) into SQL based on LookML definitions.

The SQL is sent to the connected database.

The database executes the query and returns results to Looker.

Looker formats and visualizes the data in charts or tables.

Intermediate Looker Interview Questions

9. Persistent Derived Tables (PDTs)

Q: What are Persistent Derived Tables (PDTs) in Looker?
A:

• A Persistent Derived Table (PDT) is a query result stored in the database instead of being recomputed every time.

• Looker creates a table in your database from a SQL query, and it refreshes based on a schedule.

• PDTs improve performance for complex queries by avoiding repeated computation.

Example Use Case:

• If you have a query that joins multiple large tables and applies heavy calculations, you can store it as a PDT and reuse it.

Key Points:

• Defined in LookML under derived_table with persist_for or datagroup_trigger.

• Requires write access to the database schema.

10. The extends Keyword in LookML

Q: What does the extends keyword do in LookML?
A:

• extends allows a view or explore to inherit fields, dimensions, and measures from another view/explore.

• It supports code reusability and consistency across models.

Example:

• Here, orders_extended will have both order_id and customer_name.

11. Row-level Security in Looker

Q: How do you implement row-level security in Looker?
A:

• Row-level security ensures that users see only the data they are authorized to see.

• Implemented using:

  1. Access Filters in Explores.

  2. User Attributes to store per-user security values (e.g., region, department).

  3. Adding a WHERE clause in LookML to filter data dynamically based on user attributes.

Example:

• If a user’s user_region is East, they will only see rows where region = 'East'.

12. derived_table vs view

Q: What’s the difference between a derived_table and a view in LookML?

13. Looker Caching Mechanism

Q: How does caching work in Looker?
A:

• Looker uses query caching to store the results of SQL queries temporarily, reducing load on the database.

• Cache lifetime is controlled by datagroups and persist_for settings.

• When a query is run:

  1. Looker checks if an identical query result is in cache.

  2. If yes → returns cached result.

  3. If no → runs SQL against DB, stores result in cache, and returns it.

Key Benefits:

• Improves performance.

Reduces database query costs (especially for cloud warehouses like Snowflake or BigQuery).

Advanced Looker Interview Questions

14. How do you optimize Looker dashboard performance?

Answer:

• Use PDTs for heavy queries

• Limit dashboard tiles

• Avoid unneeded joins

• Use database indexing

15.Explain the difference between yesno and tier in LookML.

Answer:

• yesno: Defines a boolean field.

• tier: Groups numeric values into defined ranges.

16.How do you integrate Looker with external applications?

Answer:
Via Looker API or Embedded Analytics features, allowing interactive dashboards inside other applications.

17. What is the difference between model and project in Looker?

Answer:

• Project: A container for all LookML files.

• Model: A file that defines database connections, explores, and joins.

18. Explain the difference between Liquid variables and LookML parameters.

Answer:
Liquid variables enable dynamic content in LookML, while parameters define user inputs for flexible querying.

 Scenario-Based Looker Interview Questions

1. Your dashboard is taking too long to load. How do you troubleshoot?

2. A user needs access to only a specific region’s data. How do you implement it?

3. How do you migrate LookML changes from development to production?

4. How do you manage Git integration in Looker?

5. How do you schedule automated email reports in Looker?

Tips to Prepare for a Looker Interview

• Review Looker Documentation and practice LookML.

• Learn SQL thoroughly, as most Looker queries run in SQL.

• Understand BI concepts like KPIs, ETL, and data modeling.

• Practice building Explores and Dashboards.

Conclusion

Looker interviews test a blend of BI knowledge, SQL skills, and platform-specific expertise. By preparing for these Looker interview questions, you can demonstrate both your technical and problem-solving abilities, increasing your chances of landing your next BI or analytics role.