Looker
What it is
Enterprise BI platform whose core is LookML, a modelling language in which data teams define dimensions, measures and relationships once; business users then explore governed data and build dashboards, and the same semantic layer serves embedded apps, APIs and AI agents. It generates SQL and runs it live against the warehouse (notably BigQuery) rather than extracting data.
Maker and history
Per Wikipedia (secondary): founded January 2012 in Santa Cruz, California by Lloyd Tabb and Ben Porterfield; Google announced the acquisition on 2019-06-06 for USD 2.6 billion and completed it in February 2020. Part of Google Cloud (Alphabet).
Products and editions
- Looker (Google Cloud core) and Looker (original) (hosted on other clouds too), per the pricing page; billing is subscription/capacity based, quote-driven. https://cloud.google.com/looker/pricing
- Looker Studio / Data Studio is a separate, free-to-start tool: Data Studio (formerly Looker Studio); “Studio Pro” is its paid tier.
Data connectivity and modelling
SQL databases and warehouses (BigQuery, Redshift, Snowflake etc.), modelled in LookML, version-controlled in Git, with CI/CD validation (GA per Next ‘26 post). Self-service Explores can blend CSV/Excel with enterprise data (GA).
Deployment
Google-hosted SaaS (core), Looker (original) hosted instances, embedded via API/iframe.
AI features (as of 2026-10-05; Next ‘26 post is dated 2026-04-22)
- Conversational Analytics in Looker, grounded in LookML: Conversational Analytics API reportedly GA for BigQuery and Looker on 2026-06-23 (search result summary, not verified). Verified (golden) queries: reported GA (not verified). An observability / token-usage dashboard is reportedly GA on Looker 26.16+ (search summary, not verified).
- Next ‘26 (preview unless noted): Dashboard Agents (preview), Embedded conversational experiences (GA), Agentic workflows (preview), Visualization Assistant (GA), Expression Assistant (preview), Insight Assistant (preview), modernised Looker interface (preview), Knowledge Catalog integration (preview).
- LookML AI agent in the Looker VS Code extension (preview): translates intent into LookML.
- MCP: managed MCP server native to Looker (preview) and open-source MCP Toolbox for Databases; see Looker MCP Server.
- BigQuery link: conversational agents shared with BigQuery; Gemini features need Google Cloud/Gemini enablement per instance (admin setting; details not verified).
Strengths and weaknesses (opinion)
Strength: one governed semantic layer and code-first modelling, which Google positions for embedded analytics and as a grounding layer for AI. Weakness: needs LookML developers, quote-based pricing (per the pricing page), and possibly less self-service polish than desktop tools (Google itself is modernising the UI, preview). Peer comparisons (Qlik, Power BI) not sourced.
Self-learning
See Looker learning path.
Sources (fetched 2026-10-05)
- https://cloud.google.com/blog/products/business-intelligence/looker-updates-for-agentic-bi-at-next26
- https://en.wikipedia.org/wiki/Looker_(company)
- https://cloud.google.com/looker/pricing
- https://docs.cloud.google.com/gemini/data-agents/conversational-analytics-api/release-notes (search result, not fetched)
Open items
- GA dates of the Conversational Analytics API and 26.16 observability come from search summaries; confirm in Looker release notes.
- Edition-level availability of each AI feature not verified.