How do you verify someone you met online?
By collecting whatever identifiers they have already given you and asking how old each one is.
The instinct is to hunt for something damning, and that is the wrong shape of search. A person who has existed for thirty years leaves an enormous amount of dull residue: a handle registered in 2011, an address in four breach dumps, a profile photograph that appears nowhere else because they took it. Nobody manufactures that, because manufacturing it would take a decade.
So you are auditing for depth rather than for guilt. Ask how far back the trail goes, whether the identifiers agree with each other, and whether anything about the picture is younger than the person claims to be.
This also means you can stop early and often. Most checks end with a long, dull history that settles the question.
Why is an empty result the strongest signal?
Because it is very hard to fake a past and very easy to fake a present.
Building a convincing profile takes an afternoon. Building a handle that has been registered on eleven services since 2013, an address that appears in a 2016 breach, and a phone number attached to accounts nobody would bother inventing takes years of ordinary living. This is why fabricated identities are shallow: not because their makers are careless but because depth cannot be produced on demand.
Read the empty result carefully, though. There is a difference between sources that answered and found nothing, and sources that failed to answer. The first is evidence. The second is a broken connection, and treating it as evidence is how people end up accusing somebody real.
Absence is evidence, not proof. Plenty of legitimate people have almost no digital history: they are young, they are private, they arrived from another country, or they deleted everything after something happened to them. An empty report is a reason to go slower, not a verdict.
What should you check, and in what order?
Cheapest and most revealing first, so that most of the time you stop after two steps.
- Their photograph, reverse searched. Free, ten seconds, and it catches the laziest fakes outright.
- Their handle. The highest-yield identifier there is, because people reuse one for a decade. 71 platforms answer for it.
- Their email address. 71 sources, and the breach records are what date the address.
- Their phone number. Thinnest of the four. Confirms the line is real and little else.
Most conversations end at step one or two, in both directions: either the photograph belongs to somebody else, or the handle turns out to have a long, boring, entirely convincing history.
The order matters more than it looks. Running the phone number first is the common mistake, because it feels like the most personal thing you have, and it is the identifier that gives back the least.
A clean report changes how confident you are. It does not change meeting somewhere public, telling somebody where you are, and arranging your own way home.
What does their photograph tell you?
Whether it is theirs, which is the single most useful fact available to you and costs nothing.
Upload it to Google Lens or TinEye. If the same face appears under a different name on a modelling site, a stock library or somebody's abandoned Facebook, you are finished and you did not need us. We do not run reverse image search and we would be worse at it than the free tools if we did.
What a file does tell us is what the camera wrote into it. Where a photograph came to you as an original rather than a screenshot, our photo analysis reads the camera, the software and any coordinates left behind. Most photographs sent through a messaging app have had all of that stripped, and that is normal rather than suspicious.
What does their phone number prove?
Less than the people-search industry advertises, and this is worth saying plainly before you pay anybody for it.
A number confirms the line exists, which carrier holds it, which country and region it belongs to, and whether it is a mobile, a landline or a virtual number. That last one is the useful part: a freshly issued virtual number, in a region that does not match anything else you have been told, is a small flag.
What it will not do is give you a name. Eight sources answer for a number here, against 71 for an email address, because a number is the identifier people guard hardest and publish least. If a service promises you a full identity from a phone number alone, it is selling you a broker file of unknown age.
What do their email address and handle prove?
Most of what you actually wanted, which is why these two are worth more of your attention than the photograph.
A handle checked across 71 platforms returns where it is taken, the display names attached, and where a platform publishes it, the join date. That join date is the whole game. An address checked against 71 sources returns the accounts registered to it and the breaches containing it, and a breach from 2016 proves the address existed in 2016.
Read for agreement rather than for volume. The same display name across four unrelated services, over several years, is worth more than fifty hits from last month. A handle that matches on one platform proves the handle is taken there, not that the same person holds it.
Which signals do not mean anything?
Four that get treated as red flags and are not, and it is worth clearing them out because they generate most of the false alarms.
- A stripped photograph. Every messaging app removes metadata on upload. This is the normal state of any picture that has been sent to anybody.
- A recent account on one platform. People join things. It only matters when every identifier is recent.
- A common handle. Generic words are registered by hundreds of unrelated people and produce hits that mean nothing at all.
- Not appearing in a breach. Good security hygiene looks identical to a new identity in this one narrow respect.
The pattern that matters is several weak signals pointing the same way at once, not any single one of them on its own.
What should you do before meeting in person?
The same things you would do without any of this, because a clean report changes your confidence and not your precautions.
Meet somewhere public and busy, and arrange your own transport both ways. Tell somebody where you are going, who with, and when you expect to be back. Keep your phone charged.
None of that is novel and all of it survives a verification that came back perfect, because the people who cause harm at a first meeting are frequently exactly who they said they were.
One pattern is worth knowing before you go, because it is the one that does not look like danger. The FTC's guidance on romance scams describes a person who is warm, consistent, and always has a reason not to meet, and who eventually asks for money. Verification catches the shallow version of that. It does not catch a patient one.
If you have found something that worries you, the useful response is to not go, and to say nothing about what you found. Telling somebody how you checked them teaches them what to fix.
How to run these checks on DetectiveCheck
Three lookups, in the order set out above, and the fourth step is the one to actually think about.
- Their handle through the username lookup, reading the join dates rather than the count.
- Their address through the email lookup, reading the breach dates for the same reason.
- Their number through the phone lookup, expecting a short report.
- Read the sources that came back empty alongside the ones that answered. That comparison is the entire method.
Every finding carries the source it came from and a confidence rating, so a strong signal and a weak one do not look alike on the page. That distinction is what stops a single thin hit from reading like a conclusion.
You can see the shape of a finished report in the sample report without creating an account.
Common questions
How can I tell if someone I met online is real?
Check whether their identifiers have a history rather than hunting for something suspicious. A real person leaves a decade of dull residue: a handle registered years ago, an address in old breach dumps, accounts on services nobody would invent. An identity built last month cannot produce that, because depth takes time rather than effort.
What is the first thing I should check?
Their photograph, through a free reverse image search on Google Lens or TinEye. It takes ten seconds and catches the laziest fakes outright, because a stolen picture usually appears elsewhere under a different name. Only move on to paid checks if the photograph survives that.
Does an empty result mean they are fake?
No. It means slow down. Plenty of legitimate people have almost no digital history: they are young, they are private, they moved countries, or they deleted everything deliberately. What matters is distinguishing sources that answered and found nothing from sources that failed to answer, since only the first is actually evidence.
Can I find out someone's real name from a phone number?
Usually not. A number confirms the line exists, its carrier, its country and region, and whether it is mobile, landline or virtual. It is the thinnest of the four identifiers, with 8 sources against 71 for an email address. Any service promising a full identity from a number alone is reselling old broker data.
Will they know I looked them up?
No. The checks query public sources and the account-existence behaviour platforms expose to anybody. Nothing is sent to them: no message, no friend request, no password reset. What can reach them is what you do afterwards, such as opening a profile on a network that shows visitors, which the lookup itself never does.
Is it legal to look up someone I am talking to?
Searching publicly available sources is legal in the United States and the EU. The restrictions are on use: in the US, deciding employment, housing, credit, insurance or tenancy from a lookup falls under the Fair Credit Reporting Act and is not permitted with a consumer service. Checking your own safety before meeting somebody is not one of those uses.
What if their photo has no metadata?
That is normal and means nothing. Every messaging app and social platform strips metadata when a photograph is uploaded, so any picture that reached you through one of them has been stripped before you saw it. Missing metadata is a fact about how the file travelled, not about the person who sent it.
Should I tell them I checked?
If everything came back fine, that is your decision and many people say so openly. If something worried you enough to stop, say nothing about what you found and simply do not go. Explaining how you checked teaches somebody who is lying exactly which part of the story to repair.
Does a clean report mean it is safe to meet?
No. It means one specific worry has been reduced. Keep every precaution you would otherwise take: meet somewhere public and busy, arrange your own transport both ways, and tell somebody where you are going and when you expect to be back. People who cause harm at a first meeting are often exactly who they claimed to be.
