Looker−8 Mixed

GILD Review

GILD's call and the buying facts behind it.

From GILD research

Should we adopt Looker?

How do you budget for Gemini in Looker before Google enforces its data token meter?

30-Second TakePick Looker if every dashboard, spreadsheet pull and AI answer must come from one set of metric definitions.
Best fitData teams on BigQuery, Snowflake, Databricks, or Redshift that want governed self-serve and AI answers from one semantic layer.GILD Review
Consider another workflowIf analysts mostly work in notebooks and ad hoc code, evaluate Hex; if the company runs on Microsoft 365 and Fabric, evaluate Copilot in Power BI first.GILD Review

Drawn from 80 sources

GILD analysis

Looker Review

Choose Looker when every dashboard, spreadsheet pull, and AI answer must come from one set of metric definitions, and a data team will maintain them. Conversational Analytics composes queries from LookML, so answers follow the joins the data team defined. The cost is the modeling work and an annual license, with a data token meter on top.

Business Intelligence
AI Adoption

The decision

Best fit: Data teams on BigQuery, Snowflake, Databricks, or Redshift that want governed self-serve and AI answers from one semantic layer.
Consider another workflow: If analysts mostly work in notebooks and ad hoc code, evaluate Hex; if the company runs on Microsoft 365 and Fabric, evaluate Copilot in Power BI first.
Adoption condition: Curated Explores with clear labels and descriptions, and data agents tested with verified queries, before Conversational Analytics opens beyond the data team.

Basis: Looker documentation, pricing, and data handling terms, plus 65 public accounts from Looker users, one of them a team's measured eight-month cost of replacing part of its Looker use. Measure in a pilot: run each data agent's verified queries as a test set, and log tokens per question in System Activity.

Buying & deployment factsWhat business and product leaders should verify before rolling it out across teams.

How you buy it

No list price

Annual commitments through Google sales, with 10 Standard and 2 Developer users included.

Source 1, Google, Pricing: Pricing

Details

Google's sales team sells Looker (Google Cloud core) on annual commitments of one, two, or three years. A platform edition (Standard for teams under 50 users, Enterprise, or Embed) includes 10 Standard and 2 Developer users, and further Developer, Standard, and Viewer licenses are added on top. Each edition caps API calls a month: Standard at 1,000 query and 1,000 administrative calls, Enterprise at 100,000 and 10,000, and Embed at 500,000 and 100,000. Google publishes no list prices.

Cost drivers

60M input tokens

The monthly Conversational Analytics pool on Standard. Every query also uses warehouse compute.

Source 1, Google, Pricing: Pricing Source 3, Google, Conversational Analytics: Conversational Analytics

Details

The platform edition and user licenses set the license bill, and Looker sends every query it composes, Conversational Analytics answers included, to the connected database to run. Conversational Analytics has a monthly data token pool per instance, with 60M input and 1.2M output on Standard, 300M and 6M on Enterprise, and 1.2B and 24M on Embed. Unused tokens expire. Overage costs US$3.00 per million input and US$20.00 per million output tokens after the promotion ends, with at least 90 days' notice.

Where it runs

Live queries

Against your connected database. Gemini and the MCP server need a Looker-hosted instance.

Source 5, docs.cloud.google.com, Looker (Google Cloud core) locations: Looker (Google Cloud core) locations Source 6, Google, Choosing a hosting option for Looker (original): Choosing a hosting option for Looker (original)Source 2, Google, Gemini in Looker: Gemini in Looker Source 7, Google, Looker-managed MCP server: Looker-managed MCP server Source 3, Google, Conversational Analytics: Conversational Analytics

Details

Looker (Google Cloud core) runs in a chosen Google Cloud region. Looker (original) is Looker-hosted on Google Cloud, AWS, or Azure, or customer-hosted. Gemini in Looker and the managed MCP server require a Looker-hosted instance. Queries run live against the connected database.

What to verify

Global processing

Model calls use a global service unless EU processing applies on Looker 26.10 or later.

Source 8, Google, How Gemini in Google Cloud uses your data: How Gemini in Google Cloud uses your data Source 3, Google, Conversational Analytics: Conversational Analytics Source 2, Google, Gemini in Looker: Gemini in Looker

Details

Gemini does not train on prompts or responses. Conversational Analytics keeps data at rest in the instance's region, but model processing uses a global service unless EU localized processing applies in Looker 26.10 and later. It remains outside FedRAMP High and Medium authorization boundaries. Conversations persist until users delete them, with no admin retention setting.