Looker

Overview

Looker, now part of Google Cloud, functions as a business intelligence (BI) and data analytics platform designed to facilitate data exploration, visualization, and application development. It is built upon a unique data modeling layer called LookML, which allows developers to define data structures, relationships, and calculations once, and then reuse these definitions across various reports and dashboards. This approach aims to ensure data consistency and accuracy across an organization's analytics efforts (What is Looker, cloud.google.com).

The platform is engineered for both technical users and business analysts through its self-service BI capabilities. Business users can interact with curated data models to generate their own reports and dashboards without direct SQL knowledge, while data professionals can leverage LookML to create complex, governed data models. Looker supports connections to various analytical databases and data warehouses, including Google BigQuery, Snowflake, Amazon Redshift, and others, often operating directly on the source database without requiring data movement into a proprietary system (Connecting to a database, cloud.google.com).

Key use cases for Looker include embedded analytics, where data insights and visualizations are integrated directly into third-party applications or websites. This allows businesses to provide data-driven experiences to their customers or internal users within their existing operational tools. Another primary application is the creation of a unified data platform, where Looker serves as a single source of truth for business metrics and definitions across an enterprise (Embedded analytics, cloud.google.com). The platform emphasizes real-time data access, allowing users to query live data rather than relying on stale extracts. This characteristic positions Looker as a tool for operational analytics and data-driven decision-making processes.

Looker's architecture supports robust programmatic access through its API, allowing developers to automate tasks, integrate Looker functionalities into custom applications, and manage instances programmatically. The platform also offers Looker Studio (formerly Google Data Studio) as a free tier option for basic reporting and visualization needs, complementing the enterprise-grade Looker Platform (Looker Documentation, cloud.google.com).

Key features

• LookML Data Modeling: A proprietary language for defining data structures, metrics, and relationships, promoting data consistency and reusability across an organization (What is LookML, cloud.google.com).
• Self-Service Business Intelligence: Provides an intuitive interface for business users to explore data, create reports, and build dashboards without requiring SQL expertise.
• Embedded Analytics: Enables the integration of Looker dashboards, visualizations, and data exploration capabilities directly into other applications, portals, or websites (Embedded analytics, cloud.google.com).
• Data Exploration and Visualization: Offers a range of visualization options and interactive tools for users to analyze data and uncover insights.
• Real-time Data Access: Connects directly to various databases and data warehouses, allowing queries to run against live data without requiring data extraction or movement.
• Looker API: A RESTful API that provides programmatic access to Looker functionalities, supporting automation, custom application development, and integration into existing workflows (Looker API Reference, cloud.google.com).
• Data Governance and Security: Features include granular access controls, row-level security, and compliance certifications such as SOC 2 Type II, ISO 27001, GDPR, and HIPAA (Looker Security, cloud.google.com).
• Version Control Integration: LookML models can be managed using Git, facilitating collaborative development and version tracking.

Pricing

Looker offers custom enterprise pricing for its Looker Platform, tailored to an organization's specific needs and usage. This typically involves discussions directly with their sales team to determine the appropriate edition and feature set. A free tier is available through Looker Studio for basic reporting and visualization.

For detailed pricing information and to request a custom quote, prospective users can consult the official Looker pricing page, cloud.google.com.

Common integrations

• Google Cloud Services: Seamless integration with BigQuery, Google Cloud Storage, and other Google Cloud products (BigQuery Integration, cloud.google.com).
• Data Warehouses: Connectors for popular data warehouses such as Snowflake (Snowflake with Looker, docs.snowflake.com), Amazon Redshift, and Microsoft Azure Synapse.
• Databases: Direct connections to various databases including PostgreSQL, MySQL, and Oracle.
• Development Tools: Integration with Git for version control of LookML models.
• Embedding Frameworks: Tools and SDKs to embed Looker content into web applications using JavaScript, Python, and other languages (Embedded analytics, cloud.google.com).
• Workflow Automation: APIs allow for integration with custom applications and workflow automation tools.

Alternatives

• Tableau: A widely used BI tool known for its strong data visualization capabilities and ease of use for business analysts.
• Power BI: Microsoft's business intelligence service, offering strong integration with other Microsoft products and a focus on self-service analytics.
• ThoughtSpot: An AI-driven analytics platform that emphasizes search and AI for data exploration, providing natural language query capabilities.

Getting started

A fundamental aspect of working with Looker is defining data models using LookML. The following example demonstrates a simple LookML model for an 'orders' table. This defines a basic view with dimensions like order_id and created_date, and a measure for total_orders.

# Define a new view called 'orders'
view: orders {
  # Specify the SQL table this view maps to
  sql_table_name: `your_project.your_dataset.orders` ;;

  # Define a primary key dimension
  dimension: order_id {
    primary_key: yes
    type: number
    sql: ${TABLE}.order_id ;;
  }

  # Define a dimension for the creation date
  dimension_group: created {
    type: time
    timeframes: [raw, time, date, week, month, quarter, year]
    sql: ${TABLE}.created_at ;;
  }

  # Define a measure to count total orders
  measure: total_orders {
    type: count
    drill_fields: [order_id, created_date]
  }

  # Define a dimension for the order amount
  dimension: order_amount {
    type: number
    sql: ${TABLE}.amount ;;
  }

  # Define a measure for the total sales amount
  measure: total_sales {
    type: sum
    sql: ${order_amount} ;;
    label: "Total Sales Amount"
  }
}

To implement this, a developer would typically:

1. Access the Looker Development environment.
2. Create a new LookML project or navigate to an existing one.
3. Define a new view file (e.g., orders.view.lkml) within the project.
4. Paste the LookML code into the file.
5. Commit the changes to Git and deploy to production.

Once deployed, these definitions become available for users to build explores, dashboards, and reports within the Looker UI (LookML Overview, cloud.google.com).