SELECT

Overview

SELECT is data cloud cost observability and optimisation, and two things about it have changed enough to invalidate older descriptions: who owns it and what it covers.

It was acquired by DoiT, announced in January 2026, and now appears in its own bylines as SELECT by DoiT, with the team intact and the product continuing independently. Its own about page has not been updated to say so.

It is also no longer Snowflake-only. Databricks is generally available and dominated its 2026 output, BigQuery has been in private preview since July 2026, and Microsoft Fabric is stated intent rather than shipped.

Connection is a dedicated service user per platform with read-only access to account metadata, plus the privilege to read and change warehouse settings, and an extra-small warehouse for its own queries.

Setup speed is a published claim worth testing: up and running in under fifteen minutes, with a named customer reporting twenty.

Automated Savings is the differentiator against a pure reporting tool. It continuously adjusts warehouse settings to improve utilisation, claiming 10 to 20% of compute spend, with per-warehouse toggles and a full audit trail of what it changed.

Note the boundary against Unravel, though: it tunes warehouse configuration and surfaces query insights for humans; it does not rewrite query code or change job code.

Cost attribution runs through usage groups, managed as JSON or YAML, so ownership can be assigned to teams, departments or projects as code rather than clicked together.

Budgets are forecast-aware, alerting when a group is projected to exceed a monthly, quarterly or annual budget, alongside anomaly detection on any metric or resource.

One practical commercial note: you can pay for it with pre-purchased Snowflake credits, which sidesteps a new budget line. The published floor is $1,499 a month and everything above that is a quote.

Key features and capabilities

The same headings are used for every cost & finops entry, so two tools can be read side by side.

What spend it covers

• Snowflake and Databricks generally available, BigQuery in private preview since July 2026
• Snowflake coverage runs end to end, across ingestion, transformation, storage and compute
• Snowflake Cortex and AI function costs are fully covered, as is Databricks Genie
• Data transfer costs gained query-level attribution in September 2026
• No cloud infrastructure, SaaS or external AI provider spend; Microsoft Fabric is intent, not shipped

Attribution and allocation

• Drill-down from top cost drivers to individual query executions
• Usage groups assign cost to teams, departments or projects, managed as JSON or YAML
• Object and query tags carry through the product for attribution
• Data transfer attributed by workload, user, warehouse and other dimensions
• The words showback and chargeback are not used on the pages read

Finding savings

• Automated Savings continuously adjusts warehouse settings, claiming 10 to 20% of compute spend
• Per-warehouse toggles let teams control which warehouses are tuned automatically
• A full audit trail records the exact actions Automated Savings took
• Query-level insights for optimisation, and identification of unused tables to drop
• No automated query rewriting, and commitment or reservation advice is not published

Budgets, forecasts and alerts

• Budgets per usage group, with cost forecasting against them
• Alerts when a team is forecast to exceed a monthly, quarterly or annual budget
• Anomaly detection on any metric or resource, covering cost spikes, quality and performance
• Team management roles and in-product audit logs, both added in 2026
• Opsgenie is the only notification integration named in the documentation

Reporting and sharing

• Custom dashboards with drag-and-drop layouts, AI-assisted configuration and sharing
• Conditional formatting on tables, and per-resource pages as the primary reporting surface
• Workload drill-down from summary to individual query
• Scheduled digests to finance are not published as a named feature
• Free public artefacts are the blog, a health check app and an open-source dbt monitoring package

Integrations

• Documented integrations with dbt, Looker, Mode, Hex, Fivetran, Tableau and Opsgenie
• Custom workloads can be tagged through query tags or comments
• A version 2 REST API covering connections, budgets, insights, actions, access and audit logs
• Usage groups configurable as code in JSON or YAML
• No orchestrator or ticketing integration is published

Where it runs and what it costs

• Software as a service only, with no self-hosted or in-network option published
• Metadata is staged out of your account to a dedicated per-customer storage bucket
• Read-only on metadata, plus the privilege needed to change warehouse settings
• It does not access customer data directly
• SOC 2 status, region choice and retention are not published on the pages read

Pricing

SubscriptionFrom $1,499 a month

One published figure, a floor of $1,499 a month, with everything above it quoted: pricing depends on factors specific to your environment, so a conversation is required. No plan tiers are published, despite the request form being headed as a custom quote. There is no free tier. A start-free-trial button appears in several places but no trial length is published anywhere. One genuinely useful commercial term: you can pay using pre-purchased Snowflake credits, so it need not come out of a new budget.

Vendor pricing page →

Demos and videos

• SELECT documentation and getting started Tutorial · select.dev
• SELECT product features Product demo · select.dev

About SELECT, part of DoiT

SELECT was founded in 2022 by Ian Whitestone, its chief executive, and Niall Woodward, its chief technology officer, drawing on their experience leading data teams at Capital One and Shopify; it publishes no headquarters. It was acquired by DoiT, announced on 7 January 2026, on undisclosed terms, with the product continuing independently and the team intact, and it publishes no funding history at all, before or after. It reported more than 250 customers over three years, naming CarGurus, Clearco, Chili Piper and Heap, with Heap citing a 40% reduction in Snowflake spend.

Founded 2022 · select.dev