🚀
FREE
+40 XP
How to Learn SEO: Sources, Evidence and Checking Advice
Learning SEO means, above all, learning to separate documentation from opinions and data from stories. SEO has plenty of conflicting advice, and the skill of checking matters more than the amount you read. Below: which sources to trust, which official Google channels exist and how they differ, how to assess studies from third-party services, a list of questions to put to any piece of advice and teaching examples from this course in which a popular claim turned out to be unsupported.
📚
Alex reads two SEO blogs
“One blog: ‘keywords in alt are the main factor’. Another: ‘we tested 10,000 pages — it hardly matters’. Before believing either, I’ll ask: what does the documentation say, how was it tested and can I check it myself?”
Levels of evidence
| Level | Source | What it gives and does not give |
|---|---|---|
| 1 | Google’s documentation (Search Essentials, Search Console help pages) | what Google says about its systems and rules; the documentation’s silence is not a refutation |
| 2 | other official Google channels: the blog, the changelog, office hours | news, explanations and answers to questions; check the date and reconcile with the documentation |
| 3 | your own data and experiments | what happened on your site; needs a control and caution (SEO experiments) |
| 4 | studies by third-party services | correlations on their own samples; causation and method need checking |
| 5 | opinions, “best practices” without data, social media | hypotheses to test, not facts |
This is not a “gold, silver, bronze” ranking by brand names: a study from a third-party service does not become reliable because the service is well known. Reliability is determined by the method and whether it can be checked.
Google’s official channels
- Search Central documentation. The foundation: Search Essentials describe technical requirements, spam policies and key best practices, and stress that meeting the requirements does not guarantee crawling, indexing and serving.
- The Search Central Blog is where official core algorithm updates, announcements of new features and SEO best practices are published.
- The documentation changelog. Dated entries with “What” and “Why” fields and an RSS feed. For example, on 1 October 2026 the guide to using generative AI content was updated with information from the Search Quality Raters guidelines; on 28 August 2026 the favicon documentation was clarified by listing the supported formats, although the formats themselves had not changed. This shows the documentation is alive, and a claim is useful to check against the date.
- The Search Status Dashboard and the list of ranking updates — to check whether your change coincided with an update (Google Updates, core updates).
- SEO Office Hours. Googlers from the Search Quality team answer questions submitted by site owners; episodes are published on YouTube about monthly.
- The Search Central Community. Help from peers and experts; participants’ answers are not official Google statements, so check what they rest on.
Remarks by individual employees on social media are useful as hints, but they are informal comments: check them against the documentation.
How to assess a study from a third-party service
- Who made it and why? A service selling a tool is interested in conclusions about the importance of its own metrics.
- What was measured? A correlation (pages in the top more often have X) does not prove that X caused the position.
- Sample and method: how many pages and which, how the data was gathered, whether there is a control, whether the data is published.
- Does it agree with the documentation? If it contradicts it, very strong reasons are needed.
- Can you check it yourself (an experiment) and does it apply to your niche.
Teaching examples from the course
| Claim | What the check showed |
|---|---|
| “DR or DA is a Google ranking factor” | these are third-party services’ metrics; Moz writes that DA is not a Google factor (profile metrics) |
| “Every page within four clicks” | there is no such rule in the documentation (site architecture) |
| “FAQPage helps voice answers” | the link is not documented, and the FAQ rich result is deprecated (voice search) |
| “Alt no longer than 125 characters” | the image guide has no such limit; it asks for useful text without keyword stuffing |
How to keep up with changes
- Subscribe to the RSS of the documentation changelog and the blog: you learn about edits on the day they are published.
- If in doubt, check the English version: wording may differ in a translation.
- Keep a change log for your own site: date, what changed, where; then you can match edits with Google updates.
- For every new piece of advice, record: the source, the date, whether it is documented, whether you have checked it.
A learning plan
- Foundation: Google’s Search Essentials and the starter guide.
- Depth: topics by task — indexing, structured data, international SEO, speed.
- Practice: your own site, Search Console and analytics, experiments (tools).
- New things: follow new features and their limits, for example AI in search (AI Overviews).
Checklist
- For each piece of advice the source and date are known.
- It is reconciled with Google’s documentation; it is stated whether it is documented or not.
- Studies by third-party services are assessed by method, not brand.
- Important claims are checked on your own data or by an experiment.
- The RSS of the changelog and the blog is subscribed to.
- A log of your own changes is kept.
Mistakes
- Believing advice because a famous blog published it.
- Treating correlation as a cause.
- Treating the documentation’s silence as a refutation or confirmation.
- Reading retellings instead of the original and ignoring the date.
- Applying advice without checking it on your own site.
What next
Next comes the module on the future of search: we start with AI Overviews, where it is especially important to tell documentation from speculation.
Lesson Task
Test your knowledge and earn +20 XP