Hierarchy For SEO
Sometimes we try to prioritize the 200 relevant criteria that Google’s algorithm takes into account. However, this is not logical and achievable, simply because it is (no longer) how it works …
We often read or hear conversations about a possible hierarchy of the 200 relevancy criteria that Google says it uses in its relevancy algorithm.
Which one has the most weight? Is that guy more important than someone else? Can we compare the importance of Pagerank with that of the title tag? Etc.
In fact, since the arrival of “ machine learning ” to engines, the situation has fundamentally changed and this notion of comparison, of “ hierarchy ”, no longer applies. And my vision is that first of all we must classify these criteria into three large families, without any real hierarchy within each one.
200 relevance criteria …
It is true that Google communicates over 200 relevancy criteria to assign a relevancy score to a web page resulting in a search being requested. This is a number that you have given for a long time and of course, there are some people who have tried to list these 200 criteria: the title tag, the H1 tag, the text volume, the PageRank, the presence of keywords in the URL, bold words, WebPerf, page load time, mobile compatibility, link anchor, backlinks so there are really a lot of criteria and the last time I saw Twitter a conversation between several people who said: is the title more important than the H1, or is the PageRank more important than the keyword in the URL, etc.
In fact, I don’t think it’s the right way to look at the problem, to ask the question.
Rank the 200 relevance criteria?
It is a fact that prioritizing the 200 relevance criteria seems totally impossible today, simply because that is not how it works.
Machine Learning
Today’s search engines work with machine learning systems, machine learning, which will manage, classify, take into account the different criteria according to many possibilities and according to a context.
Context criteria
Finally, depending on the search intention, for example, if the search is transactional, navigational or informational. Possibly based on the search history in the browser, even if it has less weight today, depending on the ” heat of the search ” ( a search that is in high demand, a news query or a ” cold request “, a linear search in the time ), a polysemy problem, with words that have several meanings such as Python, Jaguar or Salsa, depending on the geolocation, a new search. Do not forget that every day Google discovers 15% of new searches that had never been dealt with before which has no history, so there is no information to try to answer them, etc.
And so there are a lot of different contexts. And depending on these contexts, the machine learning algorithm will try to give weight to each of these criteria. But there is not, as might have been the case 15 or 20 years ago in the old search engines, there is no established hierarchy of criteria.
It all depends entirely on the context of the search intent. Therefore, it is not possible to prioritize SEO criteria in relation to others, because that is simply not how it works today.
3 major families of criteria
Crawl / Indexation
On the other hand, it is possible, (in any case, it is my vision), to define three large families of SEO criteria or points to take into account, for example, these are not criteria of relevance, that is not why we will be well-positioned, but it is these criteria that make the site well crawled, and therefore well-indexed, which is paramount in terms of SEO today.
Therefore, the canonical tag, the meta robots, the robots.txt file, the Hreflang tag for International SEO, the pagination if we have paginated content, the sitemap.xml, the no-follow links for the internal structure, etc. are absolutely important to make sure the site is well analyzed, well understood, well crawled, well indexed by the engine. It is absolutely necessary, indispensable.
Fundamental criteria
We will have fundamental criteria, those that have finally been there for 15 years, such as the title, the H1, the volume of text sufficient for the engine, so that the algorithm can understand what we want to talk about, the quality of the text, of course, the meta description, which is not a criterion of relevance, but it is something that will play in the click rate in the search results, the backlinks (PageRank), the anchoring of the links, all this is well known today.
There is no hierarchy within these criteria.
Depending on the context (which we already saw), machine learning, machine learning, the algorithm will say:
Depending on this context, it may be the title that will be very important or perhaps the volume of text or another, finally the important thing is that these fundamental criteria, all have to be on your site. Then we will let the algorithm do its “hack”, do its job, but it is absolutely necessary to have these 100% fundamental criteria. Also, I will certainly do other articles about it, because there is a lot to say about it.
The “light” criteria
We will have “light criteria”, such as mobile compatibility, web perf – the loading time of the pages or the Core Web Vitals soon, the alt attributes in the images, the presence of the keywords in the URL, in the name of the domain, in short, the things that we have to have, sometimes it is more UX than SEO, but it is better if it does SEO well, these are criteria that are not fundamental, but that we must have anyway as much as possible to put all the possibilities on our side to be well-positioned.
It is rather, these three big families that must be taken into account: of course, the mandatory crawling and indexing, the mandatory fundamental criteria without hierarchy, and the light criteria also without hierarchy, the more we will have and the better it will be. I will certainly make new articles to specify each of their families of SEO criteria, it seems quite interesting …
Latest comments