Most guides to this subject are lists of names. That format fails for a simple reason: names are unlimited and your memory is not. Sorted by structure the field collapses to five shapes, and every shape ends at the same place — you handing over control of something.
This piece describes structures and characteristics only. It does not identify or make findings about any specific company, platform or individual — that would be a legal determination, and it is not something a web page can make.
Why sorting by shape works better
A new label appears every few months for what is mechanically the same sequence. Learn ten names and the eleventh gets you, because you were pattern-matching on vocabulary. Learn five shapes and the new label is recognisable on arrival, because what you were matching on was the sequence of moves.
There is a second advantage that matters more in the moment. Names tell you what happened to other people. Shapes tell you what happens next to you, and being able to predict the next step is the most convincing thing there is. When the prediction comes true, no further argument is needed.
Shape 1 · Relationship: the long approach
Defining feature: an extended period during which money is never mentioned. That period is the product. Everything spent on it is recovered later.
The sequence is stable. Contact arrives sideways — a wrong number, a hobby group, a dating app, a professional network. Then comes a stretch of ordinary conversation, often weeks, with no pitch of any kind. Investment surfaces incidentally, as something the other person happens to do, never as a proposal to you. Interest is allowed to come from your side, which is the point of the whole preceding phase.
Then a demonstration: a modest position, a visible gain, and — the critical mechanism — a small withdrawal that completes without friction. After that, scale. Then, at the moment of a larger withdrawal, an obstacle appears.
The small successful withdrawal is the machinery
It is not an oversight and it is not evidence of legitimacy. It is a cheap purchase of your confidence, and it is the specific moment at which most people decide to commit seriously. Treat an easy small withdrawal as carrying no information about whether the larger one will work. The stage-by-stage breakdown is in a stranger offers to make you money.
Shape 2 · Impersonation: borrowing an identity you already trust
Defining feature: it does not defeat your judgement — it prevents your judgement from starting. You are not assessing a stranger; you are responding to your bank, the exchange's support desk, or a friend.
Three variants cover most of it. Official-body impersonation, where the pretext is a problem with your account. Support impersonation, which reaches you after you have publicly asked for help somewhere — a path most people never notice, covered in how fake support finds you. And personal impersonation, either a compromised account or a lookalike, which is its own file.
The reliable defence here is directional rather than analytical. Legitimate organisations do not initiate contact with you through private messages, and they do not ask for one-time codes. You do not have to work out whether the person is real. You only have to notice which direction the contact travelled in.
Shape 3 · Approval theft: nothing is transferred, and everything is lost
Defining feature: no funds move at the moment of the attack. You approve something, your balance is unchanged, and there is no sense of loss at all. That is precisely why it works.
The distinction that carries this whole shape: a transfer moves an asset once, whereas an approval grants permission to move assets later, repeatedly, at a time the other party chooses. The interfaces look nearly identical, and the approval is frequently the one that costs nothing to sign.
"It didn't cost gas, so it wasn't risky" is the most expensive misunderstanding in this category. An off-chain signature can constitute a valid authorisation that the other side submits themselves whenever they like. Mechanics, what to look at in the popup, and how to find and revoke existing approvals are in what that button actually signed.
Shape 4 · Yield promise: certainty is the tell
Defining feature: an uncertain thing described as certain. Not a high number — a definite one.
This is worth separating carefully, because most people screen on the size of the return. By that filter, a promised 8% looks conservative next to a promised 80%. But if both are described as guaranteed, they have the same structural problem; the conservative one simply takes longer to unravel, which gives more people time to put more in.
Any genuine source of return fluctuates. Lending income moves with demand, market-making income moves with volume, staking rewards move with network conditions. Fluctuation is the evidence that an external source exists. A flat, guaranteed number implies either that nobody has thought about where it comes from, or that it does not come from outside at all — in which case it comes from later depositors. The three shared tells, plus a set of questions you can put directly to a platform, are in three tells shared by every high-yield platform.
Regulators in several jurisdictions publish standing warning lists for exactly this category — the FCA warning list, the MAS investor alert list and the SFC alert list among them. A name absent from those lists proves nothing; a name present on one settles the question quickly.
Shape 5 · Gate: pay to receive what is already yours
Defining feature: to get something out, you must first put something in. Tax, deposit, unlock fee, activation, verification, account upgrade, expedite charge.
Structurally this is the simplest of the five and the easiest to call, because the paying step is itself the conclusion — you do not need to evaluate anything that comes before it. It also has no natural end: each payment produces a new reason, because each payment demonstrates that you are still willing to pay.
It appears in two positions. As the closing stage of another structure, when withdrawals suddenly require a fee. And, more predatory, as a standalone approach to people who have already lost money — the recovery offer. Anyone who contacts you after a loss promising to get funds back for an upfront fee is running this shape, and they are working from a list of people who have proven they will pay.
How they actually appear: chained, not alone
Real cases are rarely one shape. Two or three get joined, each covering a segment. Seeing the joins is what lets you predict the next segment while you are still in the current one.
| Chain | How it runs | Where it ends |
|---|---|---|
| Relationship → Yield → Gate | Trust is built, a certain return is presented, scale increases | Withdrawal blocked behind a fee |
| Impersonation → Approval | An official identity routes you to a page | One signature, drained later |
| Gate → Impersonation | After a loss, a professional-sounding party appears | A second payment, for recovery that never comes |
| Impersonation → Gate | Support says the account is restricted | A payment to "release" your own funds |
Worth noting why chaining is so common: each structure is good at a different stage, and none is good at all of them. Relationship builds trust but never asks for anything; gate extracts money but cannot establish credibility on its own. Joining them lets each cover the other's weakness, which is why the joins are where the shape becomes visible — the moment the tone shifts from one segment to the next is usually the clearest signal you will get.
Why the same person can be caught twice
The most counterintuitive fact in this subject: having been through one of these makes you more likely to be approached again, not less. It is worth understanding, because the assumption that experience protects you is itself part of what the second approach relies on.
The mechanism is unglamorous. Someone who has lost money has demonstrated three things that are hard to establish otherwise: they hold or held crypto assets, they can be persuaded to move funds, and they now have a strong motive to act. That combination is more valuable than any list of names could be, which is why the recovery offer follows a loss so reliably and so quickly.
What makes it land is that the second approach does not resemble the first. The first was an opportunity; the second is a rescue. You are not being sold anything, you are being helped, and the caution built by the first experience is pointed at the wrong shape entirely. Someone who has learned "do not trust strangers offering returns" has learned nothing about strangers offering repair.
There is a second version that catches people even without a loss. Having recognised and avoided one structure, it is natural to conclude you can recognise them generally. But the five shapes are not variations on one theme; they exploit different things. Spotting a yield promise gives you no particular advantage against an approval request, because one asks you to believe a number and the other asks you to click a button.
The practical consequence
Treat every approach as unrelated to the last one. A prior escape is not a skill you now possess, and a prior loss is not protection paid for in advance — it is an entry on somebody's list.
Concretely, that means the four tells below apply exactly as strongly on the second occasion, and the recovery offer after a loss should be treated as the most likely thing you will encounter, not the least.
The four tells all five share
If you keep nothing else from this page, keep these. They hold when the surface story is completely unfamiliar.
Four characteristics, none requiring you to judge a person
- It moves you off the official channel. Into a private message, another app, an invite-only space, a link you were given. Every shape passes through this, because none can operate where records exist and others can see.
- It describes an uncertain outcome as certain. Guaranteed, safe, fixed, protected. The certainty is the defect, independent of the number.
- It compresses your time. Deadlines, limited slots, closing today. The purpose is not urgency — it is to remove the interval in which you would check.
- It ends with you handing over control. Funds, a code, a seed phrase, an approval. Every path converges here, which is why this is the last line of defence and also the most reliable one.
Notice what none of those require: you do not have to determine who the other party is, whether their story holds together, or how professional they seem. All four are observations about the shape of the interaction, and shape is much harder to fake than manner.
To check a specific situation against these points one at a time, the suspicious platform decision tree turns them into nine questions. To match particular phrases against the five shapes, use the script matcher. Both run entirely in your browser and send nothing anywhere.
A closing observation about all four structures. None of them depends on the target being careless — they depend on the target being in a particular position, at a particular moment, with a plausible reason to act. That is why "I would never fall for that" is a weaker defence than it feels.
What the structural view offers instead is something you can actually use: recognising the shape does not require knowing the specific story. A new script arrives every few months, and every one of them still has to build trust, apply pressure, and close a gate. Watch for those three functions rather than for the details wrapped around them.
Common questions
Why sort scams by structure instead of by name?
Because names are unlimited and structures are not. A new label appears every few months for what is mechanically the same sequence. If you learn ten names, the eleventh gets you. If you learn five shapes, the new label is just a repaint of one you already recognise.
What is the single most reliable warning sign?
Being moved off the official channel. Into a private message, onto another app, into an invite-only space, onto a link someone gave you. Every one of the five structures passes through this step, because none of them can operate inside a channel where records are kept and other people can see what is happening.
I withdrew a small amount successfully. Doesn't that prove it works?
No, and this is the most expensive misreading in the whole subject. A small successful withdrawal is a standard step, not an accident. It costs the other side very little and it is the specific moment most people decide to commit more. Treat an easy small withdrawal as neutral information.
Can several of these structures appear at once?
Almost always. Real cases are usually two or three chained together, each covering one segment. The most common chain is relationship, then yield promise, then gate. Recognising the joins lets you predict the next segment before you are inside it.
Nothing on this list matches what happened to me. Does that mean it was legitimate?
No. These five cover common shapes, not every possible one. The four shared tells at the end of this piece are the more durable test: off the official channel, certainty about outcomes, time pressure, and being asked to hand over control. Those hold even when the surface story is unfamiliar.