Search algorithm criteria: relevance and positioning

How are search results determined by engines, in particular Google, Yahoo! And Bing? In fact, SERPs (Search Engine Reset Pages) are often misunderstood, misinterpreted, and rushed to conclusions based on partial elements.

This study, based on material provided by Google and analyzed by experts, describes all factors and places them in two categories: relevance and positioning. Many webmasters concentrate all their efforts on positioning and leave the main thing, relevance. But the results are determined in two stages:

  1. Pages are first selected according to their relevance for search.
  2. The selected results are then positioned by page score.
    What score includes PageRank as well as other criteria.

This hierarchy of criteria explains that a page with a high PageRank can be placed after other weaker PR pages in the SERP.

Relevance criteria

At the first stage, the engines select pages that respond to the search for an Internet user, so they contain these keywords.

Content Keywords

The keywords contained on the page are used to determine the relevance to the query made by the user. Some words have more weight than others: those that are placed in the <h1>, <h2>, etc. header tags, which are highlighted with the "strong" tag, and those that are at the beginning of the page.

Anchors

A binding is a link text on a page. The keywords contained in the binding are used to determine the relevance of the page as well as the keywords contained on the page.

Anchors often provide more accurate descriptions of web pages than the pages themselves.
Ref: The Anatomy of a Large-Scale Hypertextual Web Search Engine.

Language

The language of the page is an important criterion, pages in French are first selected by Google.fr, but if there are no results, pages in English may appear.
TLD, domain name extension contributes to the creation of a site for finding Internet users in this country. This is clarified in Google's responses to webmasters.

Geolocation

The fact that the domain name is associated with an IP address in a particular country improves positioning due to proximity, which mainly plays on sales of goods. Some hosters offer domain name geolocation for professional hosting.

Domains, directories, and file names

The domain name and full file path keywords are used to assess the relevance of a page as much as its content and bindings. The site's theme presumably depends on its domain name.
Perhaps Goggle places too much importance on the domain name. For example, if you enter BlogSmith in the search bar, you have the right to a blank page on the site blogsmith.com. You need to go to the second page of results to find out what kind of site it is .

External relations

Do I need to reduce the number of external links to transfer PR to internal pages?
This is not a good idea, primarily because external links are involved in the content, so the relevance of the page and this goes past PR, which is only a criterion for positioning. It is easy to find examples of pages with a large number of external links and which are leading in search results.
Therefore, it is necessary not to hesitate to install numerous external links, but to avoid any form of exchange that undermines the credibility of your site. Links to bad sites also devalue the page, for both engines and users.
Of course, external links must be avoided on the home page in order to save visitors opening the site.

Positioning criteria

After selecting a certain number of pages as a response to a user request, the engines try to classify the results as much as possible. This rating is obviously important, since most Internet users are limited to viewing the first results and often only the first page of results.

For this positioning phase, the engines use popularity and confidence criteria.

Trust, TrustRank

Some sites are considered trusted and enjoy greater support from search engines, moreover, minority engines that lack content in their databases than Google. You'll often see these sites appear in the results center, sometimes even with blank pages!

The link to the author's Google + profile was talked about as a trust factor, but this was denied by the MC at SMX 2014. This is not an algorithm factor.

PageRank

The PageRank for a page is assigned based on the weight of the links pointing to that page. So depending on the number of links and PR of each. This is both a criterion for popularity and trust. A popular page attributes its confidence to the pages it links.

It is often said that PageRank is no longer very important, but Google strongly discourages webmasters from the practice of monetizing links that are aimed at increasing it, at the risk of being punished, so you must believe that this criterion is always important for the engine .

Inbound links, backlinks, are positioning criteria. But outgoing external relations are the criterion of relevance. Relevance always prevails over positioning. On the one hand, they enrich content, on the other, they are used to assess the value of a page. External links to articles with good content are part of the quality of the page.

PR is the subject of controversy, errors, false interpretations that should be corrected. "PR doesn't count" is something often read on forums.

We are based on the article The Anatomy of a Large-Scale Hypertextual Web Search Engine, written by the founders of Google and the inventors of the PageRank algorithm.

The number of external references is separated by the passed PageRank.

It turns out that if the page with the starting PR 4 contains two outgoing links, we know that the weight of the PR transmitted by it is divided equally between all outgoing links. In this case, 4/2 = 2 PR units are transmitted to each of the two pages.
Ref: The Anatomy of a Large-Scale Hypertextual Web Search Engine.

The page gets a high PageRank with few backlinks.

A page can have a high PageRank if multiple pages point to it, or if multiple pages point to it and have a high PageRank.
Ref: The Anatomy of a Large-Scale Hypertextual Web Search Engine.

The quality of the links sold per page matters more than their number.

Clicks in results

This is stated in the Google PageRank patent, clicks on the results pages are taken into account.

The number of times a page is selected in search results matters, as does the time spent accessing the page.

Knowing that clicking on the result relies on the name and snipper (description), it goes without saying that they should be relevant and attractive.

Time spent on document

If the user selects a page, but immediately returns to the list of results, the engines will know about it. This page will be considered inappropriate.
Time spent on the page (unless you go back to the search results list) is ignored (Ref: MC SMX 2014).

Rebound speed

The rebound ratio is the percentage of visits that do not make the visitor look at the second page on the site.
As a quality criterion, it is very controversial, since if a visitor finds exactly what he is looking for, he does not need to look at other pages, so a high rate indicates that the page answers the question. But Google considers engagement on the site to be a more important factor, and the percentage of rebound is also given by Google communicators as a criterion for monitoring .

Panda and backlinks

Panda is the most important positioning criterion, as it can lead to a loss of 90% of traffic on the site. This mainly depends on the number of links and indexed pages returned. The ratio of the first and second should be maximum. Knowing that only one incoming link to the site is taken into account.

Penguin and Inbound Links

Starting with Penguin's algorithm, many inbound links that are considered artificial can punish a site. These are mainly links in directories or comment streams or on pages that do not have useful content.

Further information