Data Sources

Last updated: 2 May 2026

Canary is a synthesis layer over public data published by the Care Quality Commission. This page describes exactly where the data comes from, how we use it, and the boundaries we work inside. Advisors should expect to be asked these questions by clients, so we put the answers up front.

Primary source: CQC public data

All location and provider data on Canary originates from the CQC public syndication, available via:

• CQC's published datasets and APIs
• Individual inspection reports as published on cqc.org.uk
• The CQC enforcement notices register

This data is published by CQC under the Open Government Licence v3.0. Canary holds no privileged access. Anything Canary surfaces about a registered location is information CQC has already made public.

What Canary processes

Data	Source field	Refresh cadence
Location and provider directory	CQC location feed	Daily
Current and historic ratings (Overall and the five key questions)	CQC ratings feed	Daily
Inspection report metadata and PDF links	CQC inspection feed	Daily
Enforcement actions and warning notices	CQC enforcement register	Weekly
Local authority and region mapping	ONS and CQC location addresses	Quarterly

What Canary adds

The synthesis on Track and Oversee is generated by an LLM pipeline applied to aggregated views of the underlying data. Specifically, we add:

• Structure: locations grouped by LA, region, provider, and care segment.
• Trajectory: 12-month rolling rating distributions and direction of travel.
• Pattern detection: clusters of rating changes by geography, provider, and care type.
• Plain-language synthesis: monthly briefs that read the patterns and surface what an advisor would want to walk into a client meeting knowing.

Every claim in the synthesis is grounded in a counted event, a measured aggregate, or a published CQC report. We do not write commentary the underlying data does not support.

What Canary does not do

• We do not republish CQC inspection reports verbatim. We link to them at source.
• We do not modify CQC ratings. The ratings shown are exactly as published by CQC.
• We do not collect or process data about individual people receiving care, employees, or named complainants.
• We do not provide regulatory or legal advice. Synthesis is professional intelligence, not a substitute for direct engagement with CQC.
• We do not predict CQC inspection outcomes. We surface patterns. Operators and their advisors interpret them.

Refresh and lag

The location, ratings, and inspection feeds are pulled daily, typically within four hours of CQC's overnight publication window. Enforcement actions are pulled weekly, reflecting CQC's lower publication cadence on that register. The monthly synthesis email is generated on the first business day of each month using all data published up to the prior month-end.

There is an inherent lag between an inspection occurring, CQC publishing the resulting report, and Canary surfacing it. Canary does not reduce this lag. We do reduce the lag between publication and the moment you notice.

Data accuracy

The CQC public data feeds occasionally contain errors, late updates, or fields out of sync with internal CQC records. We do not correct or override these. We surface them as published. Where we identify a clear inconsistency we flag it in synthesis output rather than silently correcting it.

Licensing and attribution

CQC public data is used under the Open Government Licence v3.0. Where Canary outputs are distributed (including the monthly PDFs on Track and Oversee), the underlying data remains attributed to CQC. The synthesis layer is © Crox Ltd, licensed to subscribers under the Terms of Service.

Questions or corrections

If you spot synthesis output that doesn't tie back to the underlying data, tell us at [email protected]. We treat synthesis errors as priority bugs.