Associate Data Practitioner

Create, Modify, and Share Dashboards to Answer Business Questions

Developing Interactive Dashboards in Looker

Interactive dashboards in Looker let analysts and business users quickly explore and visualize data. A dashboard is a collection of tiles—charts, tables, text, and more—arranged on a single page. To create a new dashboard, go to the Dashboards page and click New Dashboard, then use the Add Tile option to choose from visualizations like line charts, bar charts, gauges, or tables, or to add previously saved Looks or text tiles. After adding a tile, you configure its data source, chart type, and display settings in the tile editor. Looker dashboards automatically refresh when underlying data updates, so you always see the latest information.

Organizing Content and Using Filters

Organizing content helps users focus on key insights. You can drag and drop tiles to rearrange the layout and adjust tile sizes for emphasis. Inserting text tiles creates sections or collapsible panels, keeping related charts together and making the dashboard more readable. Applying filters and dashboard parameters makes dashboards dynamic and user-friendly. You can add dashboard-wide filters that impact all tiles—for example, filtering by region or date—or tile-level filters for details on a specific chart. Parameters let users type in values to update queries on the fly, allowing viewers to tailor the dashboard to their own questions without editing the dashboard itself.

Sharing, Collaboration, and Delivery

Sharing and collaboration in Looker are straightforward. You can share dashboards with users or user groups by setting Viewer or Editor roles, or copy and fork an existing dashboard to experiment safely. Use version history to view changes, compare revisions, or revert to earlier versions. LookML-based dashboards can be managed as code, enabling teams to track and review changes in source control. For organizing content, folders store Looks, dashboards, and subfolders with access levels determining who can view or modify them. Users can use the Send or Schedule features to deliver content immediately or at recurring intervals to email addresses, webhooks, or an Amazon S3 bucket in formats like PDF, CSV, or PNG.

Visualization Techniques and Security

Data visualization in Looker uses different tools to present data in a visually appealing and informative manner. Using measures like percentages helps express key performance indicators—for example, "percent returned items" or "percent emails resulting in a sale." In LookML, percentages are computed by creating counts for conditions and linking them through formulas, such as 100.0 * ${returned_count} / NULLIF(${count}, 0). Users can drill down into their data to see the underlying details that make up aggregate numbers using sets for drill-down details. When connecting to external databases like Microsoft SQL Server, ensuring a secure connection with protocols like TLS/SSL is vital for data integrity and privacy.

Compare Looker and Looker Studio for Different Analytics Use Cases

Core Features and User Interfaces

Looker provides a structured development environment with a File Browser panel that helps users organize their work by creating folders and files within a project. Its core feature is LookML, a specialized modeling language where developers use different file types like .model.lkml and .view.lkml to define data models, relationships, and dashboards. Looker Studio has a different interface focused on ease of use, offering a drag-and-drop canvas for building reports. Users can connect directly to data files like CSVs or Google Sheets without writing code and use new chart types like Boxplot and Waterfall charts. Looker's interface is better suited for developers building governed, reusable data models, while Looker Studio's interface is better for business users who need agile, self-service reports without deep technical knowledge.

Governed Enterprise BI vs Ad-Hoc Reporting

Looker is designed for governed enterprise business intelligence using a semantic modeling layer (LookML) to create a single source of truth. Developers define dimensions, measures, and relationships in LookML files, ensuring data consistency and version control across the organization. Users explore this governed data through secure Explores without writing SQL, though Looker includes tools like SQL Runner for writing raw queries. Looker Studio is optimized for agile, self-service reporting, allowing users to connect directly to data sources and create visualizations rapidly. Users can choose between embedded data sources (tied to one report) and reusable data sources (shared across reports). Choose Looker for scalable, version-controlled data consistency where a single definition of business metrics is critical, and choose Looker Studio for rapid, independent data exploration where speed and agility are priorities.

Integration with GCP Services

Both tools integrate well with other GCP services, but their strengths differ. Looker offers deep integration through powerful APIs and supports different editions (Standard, Enterprise, Embed) for various business sizes and security needs, including features like VPC-SC. Its APIs use OAuth and other authentication methods, allowing seamless connections to services like BigQuery and Cloud Storage for complex operational workflows. Looker Studio excels in user-friendly integration with BigQuery, providing a native integration that improves query performance and offers monitoring features for real-time data processing and interactive dashboards. Within the GCP ecosystem, Looker provides robust, API-driven integration for governed, enterprise-scale analytics, while Looker Studio provides intuitive, seamless integration for agile visualization and self-service reporting.

Manipulate Simple LookML Parameters to Modify a Data Model

Defining and Applying LookML Parameters

LookML parameters are the building blocks used to model data in Looker, going into view files and model files to tell Looker how to query and present data. Dimensions describe raw data attributes that users can select and filter in an Explore, including a parameter name, a type such as string, number, or time, and a SQL expression that pulls data from the table. Optional settings include group_label for organizing fields and primary_key for unique identifiers. Measures perform calculations or aggregations on data, such as counts or sums, declared with a parameter name, a type for the aggregation method like count, sum, or average, an optional custom SQL expression, and optional filters to limit the data. Explores define starting points for queries in a model file, using parameters like explore for the name, join for the view to join, type for the join style, sql_on for the join condition, and relationship for cardinality. By combining dimensions, measures, and Explores, you create a flexible yet maintainable LookML model.

Understanding How Parameters Work Together

A LookML project consists of multiple files like model files and view files, where model files define database connections and Explore settings, forming the backbone of a project, while view files specify tables and fields focusing on how data is retrieved. Dimensions categorize data as descriptors such as Country or Product Name, while Measures provide quantifiable data like sales figures or number of clicks. By combining these with filters, you can modify how datasets are queried to produce specific reports or dashboards. Parameters use syntax like ${field_name} to refer back to defined dimensions or measures within LookML, facilitating streamlined changes across several parts if needed. Each alteration in parameters can greatly influence data queries and reports by changing how data is visually represented.

Optimizing Field Presentation and Organization

LookML allows developers to customize how data appears to end-users without changing the underlying database logic. The value_format parameter controls numeric styling using Excel-style strings, while value_format_name applies built-in formats like USD or percent, ensuring that raw numbers are presented in a business-friendly context, such as displaying 1234 as $1,234.00. The label parameter replaces the default field name with a more user-friendly title in the field picker and visualizations, while the description parameter adds a tooltip that appears when a user hovers over the field, providing essential context. The group_label parameter combines related fields into a nested list within the field picker, and the view_label parameter allows developers to move fields from one view to another in the display, creating a logical structure that groups relevant data together regardless of the underlying table structure.