Editorial Policy

This page describes how Novamya is produced. It exists so you can check whether we are following our own rules, and so that we have something concrete to be held to when we get it wrong.

What we cover, and what we do not

Novamya covers two subjects: how distribution works on social platforms, and how the market for paid social media growth operates. Anything that does not help a reader understand one of those two things does not go on the site, regardless of what it might rank for.

Specifically, we do not publish:

• Provider rankings or “best of” lists. We have a commercial relationship with two providers. A ranking produced under those conditions would be an advertisement with a scoring system bolted on.
• Reviews of individual services. Same reason.
• Pages whose only purpose is to carry a link. If a page would not be worth publishing with every outbound link removed, it does not get published.
• Near-duplicate articles. One article per idea. If two topics can only be distinguished by find-and-replace, they are one topic. This rule was on this page while we broke it: on 2 September 2026 we merged fifteen documents into nine because one argument had been restated in eleven of them with the platform name swapped, and the shingle-overlap test we relied on could not see it, because every restatement used different words. The rule stands; what changed is that we no longer trust a similarity score to enforce it.

How claims are sourced

We work from four kinds of evidence, in descending order of confidence.

1. Primary platform documentation. Telegram’s own FAQ and API documentation, Meta’s published policies and Transparency Center, YouTube’s Help Center, TikTok’s community guidelines. When a platform states something about its own behaviour, we cite the platform and link to it.
2. Published research and industry data where the methodology is disclosed and the sample is stated.
3. Direct observation. Things that can be checked by anyone: what a panel’s service catalogue contains, what a feature does when you use it, what a metric shows in a platform’s own analytics. This tier was on this page for a month before it was ever used, which we consider a failure of the policy rather than of the wording; the first of it is our survey of what fifteen panel catalogues listed on 2 September 2026, whose full dataset is published — the fifteen pages as fetched, the fetch log, all 44,064 parsed rows and the scripts — so anyone can re-check the arithmetic rather than take our word for it. The standing rules every such measurement is collected under, and the catalogue of what we have published, are on the research and data page.
4. Inference. Reasoning from the first three. This is the weakest category and it is the one that gets abused most, so we label it. When you read “appears to”, “is consistent with”, or “we cannot confirm”, that is deliberate.

Rules that follow from this:

• Numbers get a source. If an article contains a statistic, it links to where that number came from. If we cannot source a number, we do not publish the number.
• We do not invent examples. No fabricated case studies, no made-up screenshots, no invented quotes from unnamed clients.
• No fake reviews or testimonials. There are no user testimonials on this site because we do not have a way to verify them.
• We say when we do not know. Platform enforcement thresholds, algorithm weightings and internal ranking factors are not public. Where the honest answer is “nobody outside the company knows”, that is the answer we give.

Language we do not use

Some phrases are load-bearing in this industry and mean nothing. We avoid them, and when we quote them we say what they conceal.

• “100% guaranteed” — no service that depends on a third-party platform can guarantee an outcome the platform controls.
• “Instant” — usually describes queue placement, not delivery.
• “Real and active users” — an unverifiable claim about the state of an account at some future point.
• “Safe” and “undetectable” — claims about the future behaviour of a detection system the seller does not operate.
• “Lifetime guarantee” — a promise about a period longer than most providers in this market survive.

Commercial relationships

Novamya is published by Utta, which builds and maintains the websites of two social media marketing panels, SMMLaunch and AdderPanel. This is a real conflict of interest and we manage it with rules rather than with a disclaimer nobody reads.

• Links to those panels are marked rel="sponsored", which tells search engines the link is commercial.
• An article carries at most two such links. The theme enforces this in code — a third link is silently dropped rather than published.
• The link sits in a bordered box within the article, placed after a passage that describes the product rather than after one that argues against it. It is set apart from the prose deliberately, so you can tell at a glance which part of the page is commercial and which is not. It never appears in the header, the footer, the sidebar, a pop-up or a floating banner.
• Anchor text describes what is on the other end. We do not use the same keyword-stuffed anchor across dozens of pages.
• Any page carrying a commercial link displays a disclosure notice on the page itself.
• We do not accept payment from any other provider for coverage, placement, or a mention.

The plain-language version is on the disclosure page.

Who writes and who reviews

Articles are researched, written and edited by the editorial team at Utta and published under the byline “Novamya Editorial”. We do not use invented author personas.

Every article goes through the same three checks before publication: a factual pass (is every claim sourced or labelled as inference?), an editorial pass (does this answer the question a reader actually arrived with?), and a compliance pass against the rules on this page.

Those checks are real and they are demonstrably not sufficient. Eight false statements about how platforms behave, six citations attached to claims they did not support, and one invented figure presented as a research finding all passed them and reached readers; each is in the corrections log with the date it was published and the date it was caught. We describe the process because you should know what stands between a draft and this site, not because it is a guarantee.

Articles are written to a minimum of 1,200 words not because length is a virtue, but because the topics we cover cannot be answered honestly in 400 words without leaving out the part where it gets complicated. Where a subject genuinely is short, we fold it into a larger page rather than padding it into a standalone one.

Updates and dates

Every page shows when it was published and, if it has changed, when it was last updated. Both dates are real and generated from the publishing system — they are not refreshed cosmetically to make old pages look current. A third date, “reviewed”, appears only on pages someone has read back against the platform they describe; a page without one has not been re-checked since it went up, and we would rather that gap were visible than papered over.

Pages are reviewed when a platform changes something material, when a reader tells us something is wrong, and on a scheduled basis at least once a year. Where a claim was true at a specific point in time and may not be now, we say when it was checked.

Corrections

If we publish something wrong, we fix it and we say what changed. Substantive corrections — anything that would have changed a reader’s decision — are noted on the page, with the date, and listed in the corrections log. Typographical fixes are not.

The log is the test of this section. A corrections policy that has never produced a correction is a paragraph, not a practice, and for the first three weeks of this site’s life that is all it was: we shipped a real error, found it, fixed it quietly, and only wrote it down afterwards. Both the error and that failure to log it are in the log, and so is a worse one: a correction we announced as complete and had not finished, twice.

To report an error, email editor@novamya.com with the URL and what you believe is incorrect. If you are pointing at a factual claim, a source helps and will get a faster response.

How we handle uncertainty

A large share of what readers want to know about this subject is not knowable from outside the platforms. Enforcement thresholds, ranking weightings, the true composition of a purchased audience — none of these are published, and all of them are routinely stated as fact by sites competing for the same queries we do.

Our rule is to mark the boundary rather than cross it. In practice that means three registers, used deliberately:

• Stated fact, with a source. “Telegram counts a view when the post is displayed” is documented behaviour.
• Inference, flagged as such. “Completion rate appears to be the dominant signal” means it is inferred from observable behaviour and from what the platform has said, not read off a specification.
• Not known. “Nobody outside Telegram can tell you where the enforcement threshold is.” This is a complete answer and we publish it as one.

The commercial cost of this is real: a page that says “we cannot tell you” competes against pages that give a confident number, and confidence converts better. We accept that trade because the alternative is publishing invented specifics, which is the thing this site exists to argue against.

Length and depth

Articles have a floor of 1,200 words and most run well past it. That is not a search-engine target — it follows from the subject matter. The questions we cover have honest answers that are conditional, and the conditions are where the value is. A four-hundred-word answer to “does buying members get your channel banned?” has to drop either the part about what the terms actually say, the part about how enforcement is observed to work, or the part about the risk that is genuinely predictable. All three are load-bearing.

The corollary is that we do not manufacture length. Where a subject is genuinely short, it becomes a section of a larger page rather than a thin page of its own, and where two proposed articles differ only by a find-and-replace, they are one article.

Advertising and sponsored content

Novamya carries no display advertising, no sponsored posts, and no paid guest articles. We do not sell links. Unsolicited offers to buy a link, a mention or a “collaboration” are declined without reply.

Privacy and analytics

We measure aggregate traffic to understand which pages are useful. We do not sell reader data, run advertising trackers, or require an account to read anything. See the privacy policy for specifics.

Holding us to this

If you find a page on this site that breaks one of the rules above — an unsourced statistic, a third sponsored link, a “best panels” list, a page that exists only to point somewhere else — that is a bug. Report it the same way you would report a factual error, and it will be treated as one.

Related reading: about Novamya, our disclosure, and, for the same standard turned outward on the industry rather than on ourselves, our buyer’s guide to judging a panel or provider before paying it.