Sources & methodology

What we check—and what the results can actually mean.

The Tea Report brings together identity signals, available public-record context, registry checks, and credible open-web sources. This page explains how a submitted clue becomes a sourced, confidence-labeled report.

From clue to report

The method, step by step

Matching comes before reporting. That separation helps reduce the chance that a similar-name record is attached to the wrong profile.

01
Start with signals you already know
A first and last name, a US city and state, and an age or age range are required to begin. Optional phone, email, or public social-profile details can make it easier to distinguish people with similar names. These details are matching signals, not proof by themselves.
02
Find candidates before generating the full report
Candidate results bring together available identity details so you can choose the person who best matches what you know. Stronger alignment across name, age, location, and any optional signals raises match confidence. Missing or conflicting identifiers lower it.
03
Check several kinds of sources
After a candidate is selected, the report checks licensed third-party data, available public-record and court sources, published registry data, and credible public web pages. Each source category has different coverage, identifiers, and update schedules.
04
Organize findings with AI assistance
Automated systems and language models help group related items, compare signals, and turn retrieved public-web context into readable summaries. AI does not make a summary into a source, and it should not invent missing records. Review the cited link, date, and confidence label—not just the prose.
05
Show uncertainty and source context
Findings are organized by relevance and match strength. Where available, the report shows the source link, the date it was checked, and whether a detail is confirmed, likely, or only possible. Weaker items remain leads instead of being presented as settled facts.

Source landscape

Different sources answer different questions.

No single database is a complete account of a person. Using several source categories adds context, but it does not remove the gaps inside any one of them.

Identity and candidate data

What it may contribute: Names, approximate age or birth details, current or prior locations, contact signals, and public professional profiles.
Important limits: Provider coverage and update timing vary. Shared names, recycled contact details, and old locations can point to the wrong person.

Public records and court signals

What it may contribute: Available criminal or civil court, lawsuit, and property-record signals returned by licensed providers, aggregators, or public sources.
Important limits: Court systems differ by jurisdiction. Sealed, expunged, delayed, non-digitized, or otherwise unavailable records may not appear.

Published registry sources

What it may contribute: Potential matches from published US sex-offender registry data, checked against the identifying details available.
Important limits: Each jurisdiction controls what it publishes and when. A non-finding is limited to the sources and identifiers checked.

Credible open-web sources

What it may contribute: Public professional pages, social profiles, organization pages, news coverage, and other relevant public mentions.
Important limits: Web pages can change or disappear, and a page may describe someone else with the same or a similar name.

Reading results

Precise language matters.

Confirmed: Several useful identifiers align across reliable signals. The label describes match confidence, not whether a source itself is perfectly complete or current.
Likely: A strong source and supporting details align, but one or more useful identifiers are unavailable.
Possible: The item has limited support or a conflicting detail. It should be treated as a lead to verify, not a fact about the selected person.
Not found in checked sources: No matching published result was returned by the sources checked with the identifiers available at that time. It does not prove that no record or event exists.
Source link and checked date: The link lets you inspect the underlying public context where one is available. The checked date shows when the report retrieved or reviewed it; a provider’s underlying record may follow a different update schedule.

Coverage limits

Useful context still has edges.

The Tea Report currently supports searches for people in the United States only.
Sources can be incomplete, outdated, removed, delayed, or unavailable for a particular jurisdiction or person.
Common names and limited identifiers can produce unrelated candidates or ambiguous source items.
A report cannot determine that someone is safe, predict behavior, or replace your judgment and ordinary precautions.

Permitted use

Personal dating context only.

The Tea Report is not a consumer reporting agency and its reports are not consumer reports under the Fair Credit Reporting Act. Do not use a report to make decisions about employment, hiring, housing, tenants, credit, insurance, or another FCRA-regulated purpose.

Read the full informational-use notice

Privacy & accountability

Your search stays private. Source questions have a path forward.

We do not notify the person you search. Information you submit is used to run and deliver the requested report and is shared only with service providers as described in our privacy policy. If you believe a report matched the wrong person or contains inaccurate information, use the corrections process to tell us what needs review.

Read the fictional walkthrough

What we check—and what the results can actually mean.

Start with signals you already know

Find candidates before generating the full report

Check several kinds of sources

Organize findings with AI assistance

Show uncertainty and source context