Methodology

What a TrustCheck can — and cannot — tell you

A TrustCheck answers one narrow question: do the name, phone, email and address someone gave you belong to a single, real, consistent identity? This page explains exactly how that number is built, what each signal can support, and the situations where an honest person scores badly. If you only read one part, make it what it cannot know. If you are here because someone checked you, skip to if someone checks you.

Inputs

What we check

Only what you type in. TrustMatch never sees the other person’s device, profile, photos or messages, and they are not contacted or notified. Four families of signal do the work.

Phone and carrier records

What it checks:: Whether the number is a working line, what type of line it is, and how long it has been associated with the name given.
What it supports:: A number that has stayed with a major carrier for years is harder to obtain anonymously than one activated last week.
Where it stops:: Line type is a fact about the number, not its owner. Plenty of honest people use internet-based numbers — this contributes to a score, it does not decide one.

Email validity and age

What it checks:: Whether the address exists, how long it and its domain have been in use, and whether it belongs to the same identity as the rest.
What it supports:: A long-standing address that resolves to the same person as the name and phone corroborates the picture.
Where it stops:: We do not read mail, and we do not search breach or dark-web databases. A TrustCheck will never tell you an address was leaked.

Name and address consistency

What it checks:: Whether the name and address appear together in identity registries and commercially available digital-footprint data.
What it supports:: A name that has appeared at an address over years, tied to the same phone and email, is a strong corroboration.
Where it stops:: Addresses are evaluated independently of each other. Submitting several does not mean we treat one as current and the rest as history.

Usage history and consistency

What it checks:: How long the identity has been observed in the identity graph, across how many distinct years, how recently, and how steady its record has been over that period.
What it supports:: A long, consistent record is hard to fabricate. A wide swing in an identity's history, or a record that stops a year ago, is worth knowing about.
Where it stops:: This is the identity's footprint in commercial records over time — not the person's conduct, and nothing we observe directly. We never watch anyone, and a thin record most often means a young or recently-changed identity rather than a dishonest one.

Transaction risk

What it checks:: A commerce-derived indicator reflecting patterns in transaction data associated with the identity.
What it supports:: It is one input among several, and it moves the score modestly.
Where it stops:: It measures commercial history, not character. It is not a criminality signal and must never be read as one.

Output

How the score is built

The result has two halves. The Identity Score asks whether the details you submitted resolve to one person: each value is checked against identity records on its own, then against the others. The Trust Scoreasks how established and how steady that identity’s record is — how long it has been observed, across how many years, how recently, and how consistent it has been over that span. The two are weighted into the single 0–100 number.

Score	Verdict	What it means
80–100	Verified	The details corroborate each other across independent records, with history behind them.
60–79	Likely Real	Most signals line up, but at least one is thin or inconclusive.
40–59	Suspicious	Several signals disagree, or the identity has very little record behind it.
0–39	High Risk	The details largely fail to resolve to one consistent identity.

Every result ships with the individual signals behind it, so you can see which detail failed rather than only that something did. A check typically returns in about a second.

Limits

What it cannot know

This list is not hedging. These are things the product genuinely does not do, and a result that appears to tell you one of them is being over-read.

Whether someone is honest, safe, or intends to defraud you
Criminal records, sanctions lists or watchlists
Credit files, employment history or rental history
Anything about photos, video or voice — no image analysis, no deepfake detection, no liveness or biometric checks
Where anyone is — we never see the other person's IP address, and we do not geolocate. (A count of how many distinct addresses an identity has been seen at does feed the stability signal; the addresses themselves never reach us.)
Device fingerprints or browser signals — TrustMatch never runs on their device
Messages, browsing, or social media activity
Whether an email or password has appeared in a breach

A person with a perfectly consistent identity can still be a scammer. What a high score tells you is that they are probably using their own name — which is useful, and is not the same as safe.

Fairness

Why an honest person sometimes scores low

A low score means the records did not corroborate each other. There are ordinary reasons for that, and it matters that you know them before you act on a number.

Someone who recently moved

A new address takes months to appear in the records we check. The old one still looks current; the new one looks like nothing.

Someone who just changed their number

Tenure is one of the strongest signals we have, and a port or a new line resets it. A two-week-old number scores like a two-week-old number regardless of who holds it.

Someone who changed their name

Marriage, divorce and chosen names all split a record in two. The new name may have almost no history attached.

Young adults

A thin record is the normal state of being nineteen. It is not evidence of anything.

Privacy-conscious people

Someone using an internet-based number and a forwarding address is doing something sensible. Our signals cannot distinguish that from someone hiding.

People who recently arrived in the US

Our records are US-centric. A real identity established elsewhere will look sparse here.

Read the signals, not just the number. “The phone is three weeks old” and “the name does not match the address” are different problems and call for different questions.
Ask, rather than accuse.Most low scores have a dull explanation. “Is this a new number?” resolves a surprising number of them.
Let it change what you do, not what you think of someone. Meet in public, use a payment method with protection, do not send a deposit first. Those are good ideas at any score.
Treat a high score as one input. It raises confidence that the details are real. It cannot tell you what the person intends.

The other side

If someone checks you

Everything above assumes you are the one running the check. You might just as easily be the one being checked — and if you are, you are not powerless. The person checking never learns anything you did not already give them, and you can have our copy of the result removed.

You are never told who searched. We do not notify the person being checked, and we do not tell the searcher anything beyond whether the details they already had hold together.
You can ask us to suppress your details. Email us with the subject line OPT OUT and the phone number or email address you want suppressed.
We verify it is yours first. You have to control the number or address in question, so nobody can suppress someone else’s records.
Then we delete, suppress, and tell you the source. Stored results keyed to that identifier are deleted, future TrustChecks against it return no result, and we tell you which upstream data source held the information so you can use its own opt-out for records we do not control.

We respond within 45 days and charge nothing. The full detail is in the privacy policy.

Questions

Common questions

How is the TrustMatch score calculated?

Each detail you submit is checked against identity records on its own, then against the others. The score reflects two things: whether the details resolve to a single identity, and how much history sits behind that association. Phone carrier records, email validity and age, name and address consistency, and a commerce-derived transaction-risk indicator all contribute. The result is a 0–100 score, a verdict, and the individual signals behind them.

Why did a real person get a low score?

Most often because their record is thin or has recently changed — a new phone number, a recent move, a name change, being young, or having arrived in the US recently. A low score means the details did not corroborate each other in the records we can see. It does not mean the person is lying.

Does a Verified result mean the person is trustworthy?

No. Verified means the identity details match — that the person is probably using their own name. It is not a judgement about their character or intentions, and a scammer using their real details can score well. Treat the score as evidence about the details, not a verdict on the person.

What data does TrustMatch not use?

TrustMatch does not use criminal records, sanctions lists, watchlists, credit files, employment or rental history, photos, video, voice, biometrics, device or browser signals, breach or dark-web databases, or anyone's messages and social media activity. It never sees the other person's IP address or location either, though a count of how many distinct addresses an identity has been seen at over time does contribute to the stability signal.

Someone ran a TrustCheck on me. What can I do?

Email us with the subject line OPT OUT and the phone number or email address you want suppressed. We verify that you control it, delete stored results keyed to that identifier, add it to a suppression list so future TrustChecks against it return no result, and tell you which upstream data source held the information so you can use its own opt-out. We respond within 45 days and charge nothing. We never tell you who searched, and we never tell a searcher anything beyond whether the details they already had are consistent.

Is a TrustCheck ongoing monitoring?

No. A TrustCheck is a point-in-time check. Nothing is watched afterwards and you will not be notified if something changes. Running the same details again later can return a different score, because the underlying records themselves change.

Run a check and see the signals yourself

Two free TrustChecks for life, no card required. The verdict is always free.

Run a free TrustCheck