How TrustRecord data is verified

Overview

A TrustRecord is a structured, machine-readable record of operating history for a service business, issued by TrueSignal. It exists so that AI systems, search engines, and consumers can evaluate a business based on verified and sourced data rather than unsourced marketing claims.

Data on a TrustRecord comes from three tiers, each clearly labeled:

• Verified — computed by TrueSignal from authenticated system-of-record data (e.g., QuickBooks, ServiceTitan). The business cannot edit these values.
• Attested — provided by the business under a legal attestation of accuracy. TrueSignal reviews for plausibility but does not independently compute these values. Labeled as business-attested on the record.
• Public data — derived from publicly available sources (business website, Google Business Profile, state licensing boards). Labeled as public-data on the record.

The core principle: verified data cannot be changed or manipulated by the business. A business can deprecate specific data points from their record, but they cannot alter the underlying values.

Core principles

Data pipeline

The pipeline differs by data tier:

Verified data (system-of-record)

The business provides TrueSignal access to their system of record (QuickBooks, ServiceTitan, Epic EHR, etc.) — either via read-only API connection or by providing system-generated data exports. API connections are preferred because they enable automated weekly refresh.

Raw data is normalized (deduplication, job classification, geocoding, exclusion of voided/test transactions) and processed against standardized metric definitions to compute metrics.

Attested data (business-provided)

For information not available through system integrations, an authorized representative provides data under a legal attestation of accuracy. TrueSignal reviews attested data for plausibility but does not independently compute it from source systems. It is clearly labeled as “attested” on the record.

Public data (third-party sources)

Entity information, licensing status, and review signals are derived from publicly available sources. This data is collected regardless of whether the business has claimed their record. It is labeled as “public-data” on the record.

Publication

All data tiers are assembled into a TrustRecord and published in three layers simultaneously: server-rendered HTML (fully crawlable without JavaScript), Schema.org JSON-LD in the page head, and canonical JSON for direct machine ingestion. Every field carries a sourceType tag indicating its provenance.

Record status tiers

Every TrustRecord has one of three statuses. The status is machine-readable in JSON-LD and displayed on the page.

Data integrity controls

• Verified values are immutable. Metrics computed from system-of-record data cannot be edited, adjusted, or overridden by the business.
• Deprecation, not alteration. Businesses can deprecate specific data points from their public record. They cannot change the underlying values.
• Standardized time windows. Verified metrics use standardized reporting periods (trailing 12 or 24 months), applied uniformly across all businesses in a vertical.
• Source labeling is mandatory. Every data point must carry a source type (verified, attested, or public-data).
• Anomaly detection. Unusual patterns in source data are flagged for manual review before publication.

Corrections and disputes

If a business identifies an error on their TrustRecord, they can submit a correction request through the “Request correction” link on any record page. Correction requests are reviewed within 2 business days.

Corrections that can be made: factual errors in entity information, stale credential data, misattributed public data on unclaimed records.

Corrections that cannot be made: changing verified operating metrics, adjusting the time window or scope of metric calculations.

Questions

For questions about TrustRecord verification methodology, data standards, or a specific record, contact hello@usetruesignal.com.