Editorial policy

Choosing topics

We write about one thing: the specific problems ordinary people actually get stuck on, around crypto, in Kazakhstan and Central Asia.

Three tests decide whether a topic gets written: is this a situation that really occurs? Is what already exists in English adequate? Can we offer something more solid than "apparently"? All three, or we do not write it.

Explicitly not written: market analysis, coin recommendations, project reviews, anything framed as an opportunity to get rich.

Fact checking

We sort facts into three kinds and handle them differently:

The middle row is where things go wrong most easily. Plenty of guides state a fee as a precise figure, which reads as authoritative and becomes misinformation the moment the platform adjusts it — and nobody goes back to fix it. We would rather give a less impressive range than a number with an expiry date.

On first-person writing

First-person writing expresses editorial judgement and choices. Actual trades, losses and customer cases require verifiable evidence; examples without that evidence are labelled hypothetical, not presented as personal experience. Our rules:

• Judgement and choices are allowed — explain the reasons for advice. Actual experience needs supporting records; a public-page screenshot does not establish that we completed registration, a trade or a withdrawal.
• No falsely precise data. You will not find "tested at 14:23 on the 3rd" or an exact execution price — things that look like evidence and cannot be checked.
• No claims to evidence we cannot produce. We do not say "we have screenshots" without screenshots. Real screenshots of web pages are captioned with the month they were taken.

Where AI fits

We state this directly, because avoiding the question is itself a form of dishonesty.

We use AI tools in the writing process, mainly for organising source material, drafting and editing. We also hold ourselves to a few firm rules:

• Facts are not decided by AI. Anything specific about legislation, platform rules or cases goes back to a checkable source; what cannot be checked either does not get written or gets marked as uncertain.
• No fabricated evidence. No invented data, cases, user quotes or test results.
• Images are either simple geometric patterns we draw ourselves or genuine screenshots of web pages (captioned with the month). We do not use AI-generated illustration.
• A person reads it before publication, particularly the passages dealing with risk and the law.

We think readers care less about whether AI was involved than about whether the content is accurate and checkable. So the effort goes into the second — wherever something can be verified independently, we tell you how.

Writing about money and risk

• Wherever returns are mentioned, risk is mentioned alongside. No "guaranteed", "safe", "risk-free" language.
• Where loss is possible, we say so at the level of "you can lose everything you put in".
• We do not make absolute promises about things that cannot be guaranteed — preferring "usually", "roughly", "whatever the current page shows".
• For legal and tax questions that depend on your circumstances, we say to get local professional advice rather than offering an answer that looks definitive.

Corrections process

On receiving a report or finding an error ourselves: verify, fix, and log an entry on the corrections page stating what changed, why, and the date.

We do not quietly edit and pretend nothing happened. The original wording stays in the record, so you can tell whether what you read earlier was the wrong version.

Independence

This site carries exchange referral links; that arrangement and the limits we place on it are set out fully on the disclosure page. One point bears repeating here: the presence of a referral link does not change what we say about a platform's problems. We accept no paid inclusion, no sponsored content, and no editorial review by any platform.