Editorial Policy

Four principles

Our content informs decisions that cost money. A reader planning a media buying test allocates a budget based partly on what we publish about entry costs, accepted verticals and policy restrictions. That responsibility sets the standard we work to, and it is why our rules are stricter than a general publisher's would be.

Everything on this site is built on four principles. Factual accuracy means every figure traces to a named source with a verification date. Independence means no commercial relationship can alter a figure, a restriction or a named drawback. Transparency means we disclose our funding model and mark which of our sources are weaker than others. Currency means every page states when it was last checked, because in advertising policy a two-year-old fact is often simply false.

How content is produced

Step 1 — Topic selection

We choose topics by the question a reader arrives with, not by keyword volume alone. A topic qualifies when we can add something the existing results do not have: a verified figure, a primary-source policy rule, or a named discrepancy such as the deposit gap documented on our page about dating ads traffic between what a company advertises and what its documentation says. If the honest answer is that we would only be restating what is already published elsewhere, we do not write the page.

Step 2 — Primary-source collection

Verification begins with the company's own material: help center articles, developer or advertiser documentation, published terms, and official advertising policy pages. For each fact we record the source URL and the date it was read. Where a company publishes the same figure in two places with different values, we record both, because the discrepancy is itself reportable.

Step 3 — Handling gaps and conflicts

When a company does not publish a figure at all, we do not substitute a directory number and present it as verified. We either omit it or state in the body text which weaker source we used and why it is weaker. When two sources conflict, the company's own current documentation takes precedence over its blog, and both take precedence over any third party. Our full source hierarchy is documented in how we check data.

Step 4 — Drafting by the verifier

The person who did the verification writes the page. Separating research from writing is the most common way inaccuracies enter this kind of content, because a writer working from someone else's notes will fill a gap with a plausible sentence rather than going back to the source. Keeping the two roles together means every claim in the draft is one the writer personally confirmed.

Step 5 — Adversarial review

Before publication the draft is read a second time by someone taking the opposite position, with a specific checklist. Any sentence that would remain true if a competitor's name were substituted is cut or made specific. Any figure without a source note is removed. Any paragraph that could be deleted without losing an answer to a reader question is deleted. Any claim of personal testing that is not backed by actual test data is rewritten as what it really is — documented research.

Step 6 — Publication and review

Pages ship with a named author, a publication date and a verification date. We revisit a page when a platform announces a policy change, when a company we cover updates its documentation, or on a scheduled review cycle. Substantive changes are noted rather than made silently, so a returning reader can see what moved.

Fact-checking in detail

These are the source types we use, in order of authority.

Tier 1 — the company's own current documentation

Help centers, advertiser and publisher documentation, published terms and conditions, and official advertising policy pages. This is the only tier we treat as sufficient on its own for a hard figure such as a minimum deposit or a blocked-country list. We record the exact URL and read date for each fact taken from it.

Tier 2 — the company's own secondary material

Corporate blog posts, product announcements and marketing pages. These are usable when Tier 1 is silent, but they age faster and are written to sell. Where a Tier 2 claim contradicts Tier 1, Tier 1 wins and we report the contradiction rather than quietly picking one.

Tier 3 — independent documented evidence

Dated user reports on independent platforms and regulatory or court records. We use these for patterns rather than single data points, and we attribute them as reports rather than as established fact. A recurring complaint across many dated reports is evidence; one angry post is not.

Not a source — competitor articles and directories

Comparison sites and network directories are read to find what has been claimed and to identify figures worth checking, never cited as evidence. This category copies from itself, which is how a superseded figure survives for years across dozens of pages. A fact that exists only in this tier does not exist for our purposes.

Editorial independence

Editorial decisions are made on verification results alone. No company covered on this site can influence its deposit figure, its accepted-vertical status, its position in a table or the drawbacks we name about it. Commercial relationships determine which links are monetised; they do not determine what a page says.

Three specific commitments follow from that. We do not accept an arrangement conditioned on a favourable presentation, and we decline partnerships offered on that basis. We do not remove or soften an unfavourable finding at a company's request. And we do not publish scored rankings of advertising networks at all, because a score is the single field most easily bought and the least useful to a reader whose right answer depends on their own GEO, offer and budget.

Where a commercially linked company appears in a comparison, the table is built from factual columns only — founding year, minimum deposit, pricing models, formats, inventory focus — with no ratings, stars or quality ordering. Highlighting marks the commercial relationship; it does not claim superiority. The commercial terms themselves are described on the terms page.

Corrections

We publish figures that change without notice, so errors are inevitable and we would rather hear about them than not. Send the page URL and the specific figure or claim to [email protected] and, if you have one, the source that contradicts us.

1. Acknowledgement within two business days. We confirm receipt and tell you whether we are treating it as a correction, a disagreement over interpretation, or a request we cannot act on. You get an answer either way rather than silence.
2. Verification against the primary source. We go back to the company's own documentation rather than to the source you sent, because the point is to establish what is currently true, not to arbitrate between two secondary claims.
3. Correction and disclosure. Confirmed factual errors are fixed and the page's verification date is updated. Where the error was material — a wrong deposit figure, a wrong policy rule, a wrong country list — we note what changed rather than editing silently.
4. Withdrawal where needed. If a claim cannot be verified after a correction request, we remove it rather than leave it standing with a caveat. A claim we cannot support does not belong on the page in any form.

Requirements for authors

Every page on this site carries a named author with a public professional profile — currently Smith Jones, whose background and areas of focus are set out in full. We do not publish anonymously, and we do not publish under invented bylines, because an unnamed author cannot be held to anything.

Authors must have direct operational experience in paid media buying, not only writing experience — the difference shows immediately in whether a page knows which questions actually matter to a buyer. Authors must verify their own facts rather than working from a researcher's notes. Authors must disclose any commercial relationship with a company they write about, and that disclosure goes to the reader on the page, not only internally. And authors may not present research as personal testing: describing what a document says is honest, claiming a test that did not happen is not, and the distinction is enforced at review. More on the person behind this site is on the about page.