Editorial Policy
How our guides are researched, written, checked and corrected. Published so you can judge whether to trust what you read here, rather than taking our word for it.
The short version
Everything we publish is written by people who do this work for clients, checked against primary sources, and reviewed before it goes out. We do not accept sponsored posts, paid guest posts, paid links or affiliate commissions ... no company can pay to appear in a guide or to be described favourably. When we get something wrong we correct it and note the correction on the page rather than editing quietly. Guides are reviewed on a schedule, and anything about AI search is reviewed more often because the field moves fastest there.
Section 09 explains exactly how to report an error, and what we do about it.
01 Why this policy exists
Publishing openly means publishing the rules
Evariko's position is that an agency should publish the work it actually does, in enough depth that a business could run it themselves. That is a large claim, and it is worth nothing unless the content is reliable.
This page sets out the standards our published content is held to: who writes it, how claims are checked, what we refuse to do, and what happens when we are wrong. It is the document you would use to hold us to account.
A standard, not a promise of perfection
We will still get things wrong. Search changes, tools change, and some of what is confidently repeated in this industry turns out to be wrong. What this policy commits us to is a process that catches most of it and a correction habit that handles the rest visibly.
02 What we publish
The types of content this covers
Guides
Long-form explanations of how something works and how to do it, written from live client work rather than compiled from other articles.
Templates and checklists
Working documents we use internally, published in the same form rather than in a stripped-down version.
Case studies
Accounts of client engagements, published only with permission, with figures dated and their source named.
Glossary and reference
Definitions written to be usable rather than merely correct, including where a term is contested.
Service pages and commercial copy are marketing rather than editorial. They are held to the same honesty standard ... no claim we cannot evidence ... but they are not subject to the research and review process described below.
03 Who writes it
Practitioners, not content writers
Our guides are written and reviewed by people who deliver this work for clients. That is the point of them. A guide describing how to run keyword clustering is written by someone who ran it last week, not by a writer researching it for the first time.
What that means in practice
- Every guide has a named author with a bio stating their actual experience, not a generic staff byline.
- No anonymous publishing. If nobody is willing to attach their name to a piece, it is not ready.
- Where a contributor writes for us, they are identified as such and their relationship to the subject is disclosed.
- Authors write within their expertise. Someone who runs technical SEO does not write the paid social guide because a deadline needed filling.
What we will not do
We will not publish under invented author names, attribute work to a person who did not write it, or use stock photography of people who do not work here. All three are common and all three are straightforwardly dishonest.
04 How we research
Where our information comes from
Most marketing content is assembled from other marketing content, which is how a wrong claim survives a decade. We work from primary sources where they exist.
| Source type | How we use it |
|---|---|
| Official documentation | Google Search Central, Microsoft Bing, platform help centres. The first place we check for anything factual about how a system behaves. |
| Statements from engineers | Public comments from people who work on these systems, cited with date and context. Positions change, so we note when something was said. |
| Patents and research papers | Used cautiously. A filed patent is evidence of an idea, not proof it shipped, and we say so when we reference one. |
| Our own client data | Anonymised and aggregated. Described as our observation rather than as a general rule. |
| Reputable industry research | Named, dated and linked. We check the methodology before citing a figure, and skip studies that do not publish one. |
| Hands-on testing | We test tools and techniques ourselves rather than describing them from documentation alone. |
Where we could not verify something, we say it is unverified rather than phrasing it confidently and hoping nobody checks.
05 Sourcing and citation
How we handle other people's work
- Statistics are attributed and dated. Every figure names its source and the year. An undated statistic is unusable, because you cannot tell whether it still holds.
- We link to the original, not to an article that quoted it. If we cannot find the primary source, we say so or drop the claim.
- Quotes are verbatim and attributed, with enough context that the meaning is not changed by the excerpt.
- Ideas are credited. Where a framework or method originates with someone else, we name them. Reputation in this field is built by people who published first.
- Images and data are licensed or original. We do not use material we do not have the right to use.
- Competitor content is not rewritten. If the best treatment of a subject already exists elsewhere, we either add something genuinely new or link to theirs.
06 Claims and evidence
The standard a statement has to meet
Search marketing is full of claims that are repeated until they sound like consensus. We hold statements to a stated confidence level.
| Confidence | Requirement | How we phrase it |
|---|---|---|
| Confirmed | Stated in official documentation or by the platform directly | Stated as fact, with the source linked |
| Strong evidence | Consistent independent testing or widely reproduced observation | "The evidence suggests", with the evidence named |
| Our observation | What we see across our own client accounts | "In our work we see", framed as ours rather than universal |
| Contested | Credible practitioners disagree | Both positions given, and which we lean toward and why |
| Unverified | Widely repeated but without a traceable source | Flagged as unverified, or left out |
We do not publish guarantees
No guide on this site will tell you a technique guarantees a ranking, a traffic figure or a revenue outcome. Search results are decided by systems nobody outside those companies controls. Content promising certainty is either uninformed or selling something.
07 Review before publishing
What happens between draft and publication
Brief and outline
Every piece starts from a brief setting out the questions it must answer and the sources it should draw on. Writing without one is how articles end up covering the wrong subject competently.
Draft
Written by someone with direct experience of the subject, working from the brief and their own practice.
Factual check
Every claim, figure and technical instruction is verified against its source. Anything that cannot be verified is reworded to reflect its real confidence level or removed.
Practical check
Would someone following these instructions actually get the described result? Code snippets and step-by-step processes are tested rather than assumed.
Editorial check
Structure, clarity and whether the piece answers the question it promised. Padding is cut. A shorter guide that answers the question beats a longer one that circles it.
Publish and date
Published with a named author and a visible publication date, so readers can judge currency for themselves.
08 Keeping content current
Review schedules by subject
Content in this field ages unevenly. A guide to heading structure stays accurate for years; a guide to how AI assistants select sources may be wrong in six months. We review accordingly.
| Subject area | Review cycle | Why |
|---|---|---|
| AI search, AEO and GEO | Quarterly | The youngest area, changing fastest, with the least settled practice |
| Tool workflows and pricing | Every 6 months | Interfaces and pricing change without notice |
| Technical SEO | Annually | Fundamentals are stable; specific thresholds and metrics are not |
| Strategy and fundamentals | Annually | Slowest to change, but examples and figures still date |
| Anything after a core update | As needed | Major algorithm changes can invalidate advice immediately |
Updates are dated. A page shows when it was published and when it was last reviewed. We do not change the date on a page we have not actually revisited ... refreshing a date to look current while the content is untouched is misleading, and common.
09 Corrections
What we do when we are wrong
Report an error to sharia@evariko.com with the page address and what is wrong. You do not need to be a client, and you do not need to be certain.
How we handle it
- Substantive errors ... a wrong figure, a misattributed source, an instruction that would cause damage ... are corrected and a dated correction note is added to the page.
- Minor fixes ... typos, broken links, formatting ... are corrected without a note.
- Where we were wrong about something we argued, we say we were wrong rather than rewriting history so it reads as though we always held the new position.
- Acknowledgment within 2 business days, with an assessment inside 5.
Why the note matters
Silent editing is standard online and we think it is wrong. If someone acted on our guidance and it was incorrect, they should be able to see that it changed and when. A visible correction is also the only way a reader can tell whether a publisher corrects anything at all.
10 Editorial independence
What money cannot buy here
Our published content is not for sale, in any of the forms it is usually sold.
- No sponsored posts. No company can pay for an article, a mention, or favourable treatment.
- No paid guest posts. We do not accept payment to publish someone else's content, and we do not pay to place ours.
- No paid links. We neither sell links from our content nor buy them into it.
- No affiliate commissions. Tools are named because we use them. If we ever took affiliate revenue, the disclosure would appear on the page itself, not only in this policy.
- No advertising. There is no ad inventory on this site and no plans for any.
- Tool recommendations are not negotiated. No vendor has been given, or offered, influence over how they are described.
Where we recommend a tool we also say what it is bad at, which is the simplest test of whether a recommendation is independent.
11 Client work in our content
The line between publishing and confidentiality
Our guides are written from client work, which raises an obvious question about what is safe to share.
- Methods are ours to publish. How we run a topical map or a technical audit is our process, and describing it exposes nothing about any client.
- Client identity requires permission. No business is named, and no detail specific enough to identify one is used, without written consent.
- Data is anonymised and aggregated unless a client has agreed to a named case study.
- Consent is withdrawable. A client can ask us to remove a case study at any time, without explanation, and we will.
- Failures are published too, anonymised. An account of something that did not work teaches more than another success story, and publishing only wins produces a false picture of the work.
Confidentiality obligations are set out in full in our Terms of Use.
12 Frequently asked questions
Common questions about how we publish
Can I pay to be mentioned or featured in a guide?
No. Not as a sponsored post, a paid mention, a link insertion or a tool listing. The content is the reason anyone trusts the agency, and selling placements in it would spend that trust for a fee smaller than it is worth.
Do you accept guest posts?
Not as a paid arrangement. We occasionally publish a contributor with genuine expertise we lack, clearly identified as such. If you are asking in order to place a link, the answer is no.
How do I know an author actually does this work?
Every guide has a named author with a bio describing their real experience, and most link to a profile you can check independently. If a bio seems vague, ask and we will tell you exactly what that person works on.
Why do you publish methods you charge clients for?
Because knowing how something is done and having it done well are different things. Publishing the method costs us nothing we were selling, and it is the clearest evidence of competence available to someone deciding whether to hire us.
What if I disagree with something you published?
Tell us and make the case. If you are right we correct it and note the correction. If we still disagree after looking, we will say so and explain why rather than quietly ignoring it.
Do you use AI detection tools on your own content?
No, because they do not work reliably in either direction. What we rely on instead is that every claim is checked against a source by a person who understands the subject, which catches the errors that matter regardless of how a draft was produced.
How often is this policy itself reviewed?
At least annually, and whenever our process changes materially. The effective date at the top reflects the current version.
Found an error, or want to challenge something?
Send the page address and what is wrong. We would rather be corrected by a reader than keep publishing something untrue.
© Evariko, 2026. All rights reserved. This policy was last reviewed on 1 October 2026.
Read alongside our Disclaimer and Feedback Policy.